MSSQL 5 Admin

EMC NetWorker Module for Microsoft SQL Server
Release 5.0
ADMINISTRATORS GUIDE
P/N E2-2457-01
EMC Corporation Corporate Headquarters: Hopkinton, MA 01748-9103

1-508-435-1000 www.EMC.com
Copyright 1996 - 2006 EMC Corporation. All rights reserved.

Published April, 2006
EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without notice. THE INFORMATION IN THIS PUBLICATION IS PROVIDED AS IS. EMC CORPORATION MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Use, copying, and distribution of any EMC software described in this publication requires an applicable software license. For the most up-to-date listing of EMC product names, see EMC Corporation Trademarks on EMC.com. All other trademarks used herein are the property of their respective owners.
ii
EMC NetWorker Module for Microsoft SQL Server Release 5.0, Administrators Guide
Contents
Preface............................................................................................................................. ix Chapter 1 Overview

About the NetWorker Module for Microsoft SQL Server .......... 1-2 Security requirements .............................................................. 1-2 Authentication mode ............................................................... 1-3 NetWorker PowerSnap modules............................................ 1-4 Backup operations ........................................................................... 1-5 Traditional backup ................................................................... 1-5 Snapshot backups ..................................................................... 1-7 Backup levels........................................................................... 1-11 Recovery operations ...................................................................... 1-12 Traditional recovery ............................................................... 1-12 Snapshot recovery .................................................................. 1-14 Restore types ........................................................................... 1-15 Restore modes ......................................................................... 1-18 Restore time ............................................................................. 1-19 NetWorker User for SQL Server program.................................. 1-20 Fake objects.............................................................................. 1-20 Display conventions ............................................................... 1-21 Marking items ......................................................................... 1-22 Marking semantics and restrictions ..................................... 1-24 Restore window restrictions ................................................. 1-24 Error logs for backup and recovery............................................. 1-25
iii
Contents
Chapter 2
Manual Backups
About manual backups................................................................... 2-2 Performing a backup with NetWorker User for SQL Server..... 2-3 Task 1: Start the NetWorker User for SQL Server program ......................................................................... 2-4 Task 2: Select the SQL Server data ......................................... 2-6 Task 3: Set the backup options ............................................... 2-7 Task 4: Set the backup properties for each marked database..................................................................... 2-10 Task 5: Start and monitor the backup.................................. 2-11 Task 6: Back up the client indexes and bootstrap file ....... 2-12
Chapter 3
Scheduled Backups
About scheduled backups .............................................................. 3-2 Setting backup levels....................................................................... 3-3 Differences between backup levels........................................ 3-4 Configuring scheduled backups.................................................... 3-7 Task 1: Assign a Snapshot Policy (optional)......................... 3-8 Task 2: Configure one or more group resources................ 3-10 Task 3: Configure one or more Client resources................ 3-11 Task 4: Configure a schedule and set backup levels ......... 3-13 Task 5: Configure the NetWorker volume pools............... 3-17 Unavailable databases and group backup failure..................... 3-19
Chapter 4
Restoring SQL Server Data

Before the restore ............................................................................. 4-2 Rollback restore requirements................................................ 4-3 Piecemeal restore requirements ............................................. 4-4 Backing up the active portion of the transaction log.................. 4-4 Backup the transaction log ...................................................... 4-5 Backing up the transaction log for SQL Server 7.0 .............. 4-6 Backing up the transaction log for SQL Server Express ..... 4-7 NetWorker Module database restore process.............................. 4-7 Performing a restore with NetWorker User for SQL Server...... 4-8 Task 1: Set up the restore......................................................... 4-9 Task 2: Specify the browse time (optional) ......................... 4-33 Task 3: View the required volumes (optional) ................... 4-35 Task 4: Set the restore properties (optional) ....................... 4-36 Task 5: Start the restore.......................................................... 4-54
iv
Contents
Chapter 5
Backup and Recovery for Microsoft Cluster Server

How the NetWorker client works in a Microsoft Cluster Server .................................................................................. 5-2 How the NMSQL works in a Microsoft Cluster Server............. 5-3 How the NMSQL detects SQL Server instances.................. 5-3 Named instances in failover cluster configurations ........... 5-4 Active/Passive cluster configurations.................................. 5-4 Active/Active cluster configurations ................................... 5-4 How to run a scheduled backup ................................................... 5-5 Requirements for scheduled backups................................... 5-5 Sample configuration for scheduled backups ..................... 5-7 Configure scheduled backups.............................................. 5-12 How to run manual backups and restores for a clustered SQL Server...................................................................................... 5-15 Start a backup or restore from the NetWorker User for SQL Server program .............................................................. 5-15 Start a manual backup or restore from the command prompt .................................................................. 5-16
Chapter 6
Microsoft SQL Server

Microsoft SQL Server recovery models ....................................... 6-2 Full recovery model ................................................................. 6-2 Bulk_Logged recovery model ................................................ 6-2 Simple recovery model ........................................................... 6-3 Specifying database recovery models................................... 6-3 Changing the recovery model for a database ...................... 6-5 Microsoft SQL Server 2000 and 2005 named log marks ............ 6-7 Transaction log maintenance ......................................................... 6-7 How to prevent log overflow ................................................. 6-8 SQL Server master database maintenance ................................ 6-11 Database consistency check ......................................................... 6-11 Perform a database consistency check ................................ 6-12 Multiple Nonclustered Instances of SQL Server ...................... 6-13 How a restore interacts with SQL Server................................... 6-15 Restoring the SQL Server master database ........................ 6-15 Restoring the SQL Server master database in a cluster.... 6-16 Restoring the SQL Server msdb database .......................... 6-16 Restoring both the SQL Server master and msdb databases ................................................................................. 6-16
Contents
Chapter 7
Disaster Recovery
Disaster recovery features .............................................................. 7-2 System database restore automation ..................................... 7-2 Database restore order............................................................. 7-2 SQL Server startup complete detection................................. 7-3 Overwriting capability............................................................. 7-3 Procedures for disaster recovery ................................................... 7-4 When not to reinstall SQL Server........................................... 7-4 Recovery of a damaged primary disk of the SQL Server ... 7-5 Recovery of a damaged binary disk of the SQL Server ...... 7-6 Recovery of SQL Server and NetWorker server .................. 7-7 Recovery of SQL Server without reinstalling SQL Server (shortcut).................................................................................... 7-8 Recovery of SQL Server........................................................... 7-9 Use the NetWorker User for SQL Server Program to complete a disaster recovery................................................. 7-10
Appendix A
NetWorker Module Commands

Overview of the NMSQL commands ......................................... A-2 Using the nsrsqlsv command ....................................................... A-3 Using NetWorker 7.2 and earlier .......................................... A-3 Using NetWorker 7.3 and later.............................................. A-4 Command syntax for nsrsqlsv............................................... A-4 Command options for nsrsqlsv ............................................. A-5 Using the nsrsqlrc command ....................................................... A-8 Command syntax for nsrsqlrc ............................................... A-8 Command options for nsrsqlrc.............................................. A-9 Sample restore command lines............................................ A-14 Using the nwmssql command ................................................... A-15 Command options for nwmssql.......................................... A-15 Backup and restore command syntax for SQL Server data .................................................................................... A-16 Syntax for a named instance configuration for SQL 2000 and 2005 ................................................................ A-17 Syntax for names containing a period, backslash, or colon ................................................................ A-19
Appendix B
NetWorker Module Permissions Wizard

Permissions wizard features ........................................................ B-2 How to run the permissions wizard ........................................... B-2
vi
Contents
Appendix C
Striped Backup and Recovery

Striping with the NMSQL ............................................................ C-2 Interleaving .................................................................................... C-2 Striped backups ............................................................................. C-3 Performance considerations for striping ............................. C-3 Perform a striped backup ...................................................... C-4 Striped recoveries .......................................................................... C-5 Optimal striped recovery versus fail-safe striped recovery....................................................................... C-5 Performing an optimal striped recovery (best-case scenario) ................................................................. C-6 Performing a fail-safe striped recovery ............................... C-6 Windows registry entry for striped backup...................... C-10
Glossary ........................................................................................................................ g-1 Index ............................................................................................................................... i-1
vii
Contents
viii
Preface
As part of an effort to improve and enhance the performance and capabilities of its product line, EMC from time to time releases revisions of its hardware and software. Therefore, some functions described in this guide may not be supported by all revisions of the software or hardware currently in use. For the most up-to-date information on product features, refer to your product release notes. If a product does not function properly or does not function as described in this guide, go to http://softwaresupport.EMC.com. Audience This guide is part of the EMC NetWorker Module for Microsoft SQL Server (NMSQL) documentation set, and is intended for use by system administrators during installation and setup of the product. Operators who monitor daily backups may also find this guide helpful. Readers of this guide are expected to be familiar with the following:

EMC NetWorker products Microsoft SQL Server versions: 7.0, 2000, and 2005 Express
ix
Preface
Organization Chapters and appendixes in this guide are as follows:

Chapter or appendix Chapter 1, Overview Describes This chapter provides information about how the NetWorker Module for Microsoft SQL Server software backs up and restores Microsoft SQL Server data. This chapter explains how to use the NetWorker User for SQL Server program to manually back up Microsoft SQL Server databases. This chapter explains how to configure scheduled backups of SQL Server data. This chapter describes the steps required to restore Microsoft SQL Server data by using the NetWorker User for SQL Server program. This chapter explains the NetWorker Module backup and restore procedure for a Microsoft Cluster Server (MSCS) environment. This chapter addresses Microsoft SQL Server topics.
Chapter 2, Manual Backups
Chapter 3, Scheduled Backups Chapter 4, Restoring SQL Server Data Chapter 5, Backup and Recovery for Microsoft Cluster Server Chapter 6, Microsoft SQL Server
Chapter 7, Disaster Recovery This chapter provides procedures for recovering SQL Server data in the event of a disaster. Appendix A, NetWorker Module Commands Appendix B, NetWorker Module Permissions Wizard Appendix C, Striped Backup and Recovery This appendix provides syntax, notation, and related information for NetWorker Module for Microsoft SQL Server commands and command options. This appendix describes how to use the NetWorker Module Permissions Wizard. This appendix describes how the NetWorker Module uses striping to improve backup and recovery performance.
Related documentation The following EMC sources provide additional information relevant to NMSQL 5.0. Related documents include:
NMSQL 5.0 documentation set: Installation guide Release notes Online help
Preface
EMC NetWorker PowerSnap Module Installation and Administrators Guide
These sources, specific to the NetWorker server version, are also available:

EMC NetWorker Administrators Guide EMC NetWorker Installation Guide EMC NetWorker Release Notes
Refer to the SQL Server documentation from Microsoft for procedures and administrative information. Conventions used in this guide EMC uses the following conventions for notes and cautions.
Note: A note presents information that is important, but not hazard-related.
CAUTION A caution contains information essential to avoid data loss or damage to the system or equipment. The caution may apply to hardware or software. Typographical Conventions EMC uses the following type style conventions in this guide:
AVANT GARDE
Keystrokes

Palatino, bold
Dialog box, button, icon, and menu items in text Selections you can make from the user interface, including buttons, icons, options, and field names New terms or unique word usage in text Command line arguments when used in text Book titles
Palatino, italic
xi
Preface
Courier, italic Courier
Arguments used in examples of command line syntax. System prompts and displays and specific filenames or complete paths. For example:
working root directory [/user/emc]: c:\Program Files\EMC\Symapi\db
Courier, bold
User entry. For example:

symmpoll -p
Options in command line syntax
Where to get help For EMC technical support, product sales, and licensing information, go to: http://softwaresupport.EMC.com. Your comments Comments and suggestions about our product documentation are always welcome. To provide feedback: 1. Go to:
http://softwaresupport.EMC.com.
2. Click the Documentation link. 3. Click the Technical Publications Customer Survey link at the top of the page to comment on a particular guide.
xii
1
Invisible Body Tag
Overview
This chapter provides information about how the EMC NetWorker Module for Microsoft SQL Server (NMSQL) software backs up and restores Microsoft SQL Server data. This chapter includes the following sections:

About the NetWorker Module for Microsoft SQL Server ............1-2 Backup operations .............................................................................1-5 Recovery operations ........................................................................1-12 NetWorker User for SQL Server program ....................................1-20 Error logs for backup and recovery...............................................1-25
Overview
1-1
Overview
About the NetWorker Module for Microsoft SQL Server

The NetWorker Module 5.0 for Microsoft SQL Server is a NetWorker add-on module that provides backup and restore of the following database and transaction logs:

Microsoft SQL Server 2005 Microsoft SQL Server 2005 Express Microsoft SQL Server 2000 Microsoft SQL Server 7.0
The NetWorker software provides backup and restore capabilities for file system data only. A file system backup, however, does not save SQL Server data in a recoverable form. The NMSQL enables the NetWorker software to back up and restore Microsoft SQL Server data. The NetWorker Module for Microsoft SQL Server Installation Guide provides details about NetWorker software and NMSQL configuration.
Security requirements
Use of the NMSQL to back up Microsoft SQL Server databases requires that the proper privileges be granted to the NMSQL process. Microsoft SQL Server imposes the following requirements on third-party backup products, such as the NMSQL:
The logon account that the third-party backup process uses to connect to SQL Server must be granted the SQL Server system administrator (sysadmin) role in order to issue the T-SQL BACKUP query. The Windows logon account under which the third-party backup process is running must be granted the SQL Server sysadmin role in order to open a shared memory handle when initializing the Microsoft Virtual Device Interface (VDI).
NMSQL imposes the following requirements:
When using the nsrsqlsv and nsrsqlrc commands, the Windows logon account must be granted the SQL Server sysadmin role. NetWorker User for SQL Server must be a member of the local Backup Operators group.
1-2
Overview
NetWorker User for SQL Server must be a member of the local Administrators group.
CAUTION If running the NMSQL on Microsoft Windows Server 2003, user accounts or groups are not required to be a member of the local administrators group. However, if you do not have administrative privileges, then run the NMSQL software from the console, not from Remote Desktop/terminal services. Otherwise, it causes failures even if all permissions are correct.
Note: Windows Server 2003 enforces new security constraints in which the default user account is limited, and does not have the necessary privileges to run the NMSQL. The Permissions Wizard however, enables a local system administrator to easily configure user accounts and groups to run the NMSQL. Appendix B, NetWorker Module Permissions Wizard provides more information.
Authentication mode
Microsoft SQL Server provides two authentication modes:

Windows authentication mode Mixed mode (Windows authentication and SQL Server authentication)
Regardless of the authentication mode, you must add the Windows logon account to the SQL Server sysadmin role under which the NMSQL runs. Windows authentication mode When running in Windows Authentication Mode, the Windows logon account that the NMSQL process uses must be granted the SQL Server sysadmin role. The administrator and BUILTIN\administrator accounts are automatically members of the sysadmin role. Mixed mode When running in mixed mode and using a SQL Server logon account to connect to Microsoft SQL Server, the logon account must be granted the sysadmin role. Microsoft documentation provides more information on how to add members to a SQL Server role.
About the NetWorker Module for Microsoft SQL Server
1-3
Overview
NetWorker PowerSnap modules

The NMSQL software supports NetWorker PowerSnap Modules, which are interfaces between a Snapshot-capable storage subsystem and the NetWorker and NMSQL software. By using the PowerSnap Module appropriate for the SQL Server storage subsystem, you can create and manage point-in-time (PIT) copies (snapshots) of Microsoft SQL Server 2000 and 2005 data. The NMSQL supports Snapshot operations in environments where SQL components are located on hardware that comprise multiple storage subsystems, provided that each storage subsystem is Snapshot capable. The NetWorker PowerSnap Module Installation and Administrators Guide for the specific storage subsystem provides more information on PowerSnap Modules. The Compatibility Guides at http://softwaresupport.EMC.com provides a current list of supported storage subsystems. Homogenous storage platform environment The NMSQL supports Snapshot backup operations in homogenous storage platform environments only. All SQL components (databases and log files) must be located on a Snapshot-capable storage subsystem. For environments in which any of the SQL components are on hardware that is not Snapshot capable, the NMSQL supports traditional backup and restore operations only.
Note: When processing a Snapshot backup request, if the NMSQL detects that any SQL objects included in the request are located on storage hardware that is not Snapshot capable, the backup operation terminates and an error message appears.
LAN and LAN-free environments In LAN and LAN-free environments, the NMSQL supports the following:

Snapshot operations Serverless backup method
1-4
Overview
The NetWorker PowerSnap Module Installation and Administrators Guide for the specific storage subsystem provides more information about support for LAN and LAN-free environments.
Backup operations
The following sections provide an overview of the NMSQL traditional and Snapshot backup operations.
Traditional backup
You can run a traditional backup of SQL data at any time, completely independent of any scheduled backups. Chapter 2, Manual Backups provides information on manual backups.
Note: Traditional backups are often referred to as manual backups or as ad hoc backups.
To configure a scheduled backup to start at a designated time and perform a routine backup procedure, an administrator must set the NetWorker server resource attributes by using the NetWorker Administrator program or the NetWorker Configuration Wizard. Configuring scheduled backups on page 3-7 provides more information on scheduled backups. For traditional backups, Microsoft SQL Server supports database, file, filegroup, and transaction log backups. The NMSQL provides the mechanism that integrates the SQL database backup technology with the NetWorker software. Microsoft SQL Server 2000 and 2005 provide support for backing up and restoring filegroups and files. In addition to creating a level full file or filegroup backup, SQL Server 2000 and 2005 support the creation of filegroup differential and file differential backups. A filegroup differential backup may actually reduce both media requirements and restore time. The data can be stored across more than one disk or disk partition, and restore time may be reduced. A differential can substitute for any log backups performed between the full and differential backups. You must perform the full backup first.
Backup operations
1-5
Overview
Note: If a database has been made read-only, a full backup of the database should be made. A restore of the read-only database cannot then be made from a transaction log backup that may already exist. If a backup was created by using NMSQL release 3.0 or later, a SQL Server 7.0, 2000, or 2005 file or filegroup can also be restored from a full database backup.
Traditional backup process Figure 1-1 on page 1-6 shows an overview of the process interactions among the NetWorker client and server, NMSQL, and SQL server software during a traditional backup.
NetWorker Client Microsoft SQL Server Databases and Logs NetWorker Server File Index Media Database Storage Media
SQL Backup API
nsrsqlsv
NetWorker XBSA
nsrindexd nsrmmdbd data
inter-process communication
nsrmmd
nsrexecd
savegrp
nsrd
Figure 1-1
Traditional backup command and data flow
1-6
Overview
The process is explained here: 1. The nsrd program starts a traditional backup on the NetWorker server. 2. The savegrp program executes the NMSQL backup command (nsrsqlsv) on the client instead of performing a standard NetWorker save. 3. The nsrsqlsv program passes the backup data from SQL Server to the NetWorker server through an X-Open Backup Services application programming interface (XBSA). The NetWorker server software performs all scheduling and storage management tasks. The administrators guide provides information about the NetWorker services and operations described in this chapter.
Snapshot backups
Snapshot backups are configured as scheduled backups on the NetWorker server. The NMSQL does not support manual Snapshot backups from either the NetWorker User for SQL Server program or the command prompt. The NMSQL supports:

Full snapshot backups of SQL Server databases. Backups of only one database per scheduled backup.
Note: A snapshot backup fails if more than one database, or MSSQL, is specified for the Save Set attribute. For snapshot backups, list only one database for the Save Set attribute. Refer to your Microsoft SQL documentation for additional information.
The NMSQL does not support:
Differential or incremental (transaction log) backups for databases under a snapshot schedule. Snapshot backups for individual filegroups or files. Snapshot backups of the SQL Server master database.
Backup operations
1-7
Overview
Note: Store snapshots on a separate volume. Databases intended for Snapshot operations should be isolated on their own volumes. Databases may span one or more volumes for data and log files. It is not necessary to install SQL Server on a volume capable of storing snapshots, unless file system snapshots are intended to maintain SQL Server files (not databases). Other (non-database) files located on database snapshot volumes will be overwritten during rollback operations.
The NMSQL supports three types of snapshot backups:

Instant Nonpersistent Serverless
Instant backup An instant backup creates a PIT (point in time), or snapshot, of a SQL Server database and retains the snapshot on the SQL Servers primary storage subsystem. Depending on how backups are configured, a snapshot created during an instant backup may or may not be moved to secondary storage on the NetWorker server or storage node. Whether the snapshot should be retained is dependent on the snapshot policy. There are three ways to manage PIT snapshot backups:
A PIT copy of the data is created and immediately backed up to tape or disk, and the original snapshot is deleted after the backup is complete. A save set that is created on tape is called a Rollover Save Set. This process enables you to perform a rollover restore. An existing PIT copy of the data can be copied to a tape or disk, much like a traditional NetWorker backup and the original PIT copy, is retained on the SQL Servers primary storage subsystem. Thus, two copies of this backup exist. This process enables you to perform a rollover or PIT restore. A PIT copy of the data is created and retained on the SQL Servers primary storage subsystem and no other copy is maintained. You can use this copy to perform a PIT restore only once. If the PIT restore fails for any reason, you must restore from another backup (another snapshot or normal tape).
1-8
Overview
While the snapshot resides on the SQL Servers primary storage subsystem, it is referred to as a persistent snapshot. Retaining persistent snapshots on primary storage enables the NMSQL to perform an instant restore. Rollback restore type on page 1-17 provides information about rollback restores. Depending on the capabilities of SQL Servers storage subsystem, schedule instant backups to be performed several times per day. By scheduling frequent instant backups, such as every few hours, exposure to data loss is minimized. You can quickly perform an instant restore to return the SQL Server to a recent point in time. Nonpersistent backup A nonpersistent snapshot backup creates a point-in-time copy of a SQL database, and then immediately moves it to secondary storage on the NetWorker server or storage node. The original snapshot is automatically deleted from primary storage.
Note: In NMSQL log files, the operation type for a nonpersistent snapshot backup operation is conventional backup.
Serverless backups In a serverless backup, a snapshot is created on the SQL Servers primary storage subsystem and then immediately moved to secondary storage by a proxy client. There is no need for the SQL Server host to move the data to the secondary storage medium (typically tape). Snapshot backup process Figure 1-2 on page 1-10 shows an overview of the interactions among the NetWorker client and server, NMSQL, PowerSnap Module, and SQL Server software during a snapshot backup.
Backup operations
1-9
Overview
NetWorker Client Microsoft SQL 2000 Server Databases and Logs
NetWorker Server File Index Media Database
Storage Medium
nsrindexd PowerSnap nsrsqlsv BRC nsrmmdbd data nsrsnap nsrmmd
nsrexecd inter-process communication
savegrp
nsrd
Figure 1-2
Snapshot backup command and data flow
The PowerSnap Module Backup Recover Control (BRC) service, running on the SQL Server, provides Snapshot functionality to the NMSQL. Through the BRC application programming interface (API), the NMSQL determines if SQL Server databases and transaction logs are located on Snapshot-capable hardware, and therefore eligible for Snapshot backups. The BRC API also provides NetWorker indexing and media database services, and enables the NMSQL to specify which files are moved to secondary storage. The PowerSnap Modules nsrsnap program is invoked when the NetWorker server initiates a scheduled Snapshot backup. The nsrsnap program queries the NetWorker server for configuration information, such as the Snapshot policy, then executes the nsrsqlsv backup program with a command that saves a snapshot on the primary storage.
1-10
Overview
Snapshot data mover The snapshot data mover (also called proxy client) is the computer that actually moves the data during a Snapshot operation. By default, the SQL Server host is the data mover. However, you can use the NSR_DATA_MOVER attribute to specify a different computer to act as data mover. In this case, the PowerSnap nsrsnap_save program on the data mover computer performs the backup.
Backup levels
The NetWorker software provides three main levels of backup: full, incremental, and differential, where differential is specified as any level from 1 to 9. The NMSQL also provides full, incremental, and differential backups.
An incremental backup, done from the command line, corresponds to a Microsoft SQL Server transaction log backup. A log file backup by itself cannot be used to restore a database. A log file is used after a database restore to recover the database to the point of the original failure. A differential backup, specified as any level from 1 to 9, is done from the command line, and makes a copy of all the pages in a database modified after the last full database backup.
Table 1-1 on page 1-11 shows how the terminology for backup levels used in the Microsoft SQL Server product differs from the terminology used in the NMSQL product.
Table 1-1
Terminology Function Backs up an entire file, filegroup, or database. Backs up all transaction log changes since the most recent full, differential, or transaction log backup. Backs up all database changes, including filegroups and files, since the last full backup. NMSQL term Full file, filegroup, or database backup (full) Incremental database backup (incr) File, filegroup, or database differential backup (diff) SQL Server term File, filegroup, or database backup Transaction log (also called xlog) backup Differential backup
Backup operations
1-11
Overview
Figure 1-2 on page 1-10 summarizes where the NMSQL backup procedures can be initiated and which backup levels are supported for each interface.
Table 1-2
Where to initiate backup operations Backup levels available
Backup type Scheduled Manual
Backup initiated from NetWorker Administrator program on the server Command line on the NetWorker server that is a client host NetWorker User for SQL Server program on the client
full Yes Yes Yes
incr Yes Yes No
diff Yes Yes No
Recovery operations
You can recover data from a traditional or Snapshot backup at any time by using the NetWorker User for SQL Server program. Alternatively, you can recover data from a traditional backup by running the NMSQL recover command (nsrsqlrc) from the command prompt. You cannot recover snapshot backups from the command prompt.
Traditional recovery
Using NMSQLs traditional recovery operation, you can recover files, filegroups, databases, and transaction log backups. Chapter 4, Restoring SQL Server Dataprovides additional information about traditional recovery operations. Traditional recovery process Figure 1-3 on page 1-13 shows the functional relationship between the NetWorker server, the NMSQL, and the SQL Server products during a traditional recovery operation.
1-12
Overview
NetWorker Client Microsoft SQL Server Databases and Logs
NetWorker Server File Index Media Database
Storage Media
SQL Backup API
nsrsqlrc
NetWorker XBSA
nsrindexd nsrmmdbd data nsrmmd
inter-process communication nsrd
Figure 1-3
Traditional recovery command and data flow
A request for a traditional restore operation: 1. The nsrsqlrc command is invoked for the NMSQL recover. 2. The NetWorker XBSA API translates the object names requested by the NMSQL into a format the NetWorker software understands, and forwards them to the NetWorker server nsrd service. 3. The media service, nsrmmd, invokes nsrmmdbd to search the NetWorker servers media database for the volumes that contain the objects requested. 4. After the media is mounted, the nsrmmd program sends the data through the NetWorker XBSA API to nsrsqlrc, which recovers the data to the Microsoft SQL Server directories.
Recovery operations
1-13
Overview
Snapshot recovery
A snapshot recovery operation can be performed at the file, filegroup, or database level from a full database snapshot. The NMSQL supports one type of snapshot restore operation called an instant restore. An instant restore operation recovers data from a PIT snapshot, but does not eradicate the original snapshot. Chapter 4, Restoring SQL Server Data provides additional information on snapshot recovery.
Note: The NMSQL does not support a snapshot recovery of the SQL Server master database.
Snapshot recovery process Figure 1-4 on page 1-14 shows the interaction among the NetWorker client and server, NMSQL, PowerSnap Module, and Microsoft SQL Server software during a snapshot recovery operation.
NetWorker Client Microsoft SQL Server Databases and Logs NetWorker Server File Index Media Database Storage Medium
nsrindexd PowerSnap nsrsqlrc BRC nsrmmdbd data nsrmmd inter-process communication
nsrd
Figure 1-4
Snapshot recovery command and data flow
1-14
Overview
A request for a snapshot recovery: 1. The nsrsqlrc command is invoked for the NMSQL. Snapshot recoveries are managed by the PowerSnap Backup Recovery Control (BRC) service, through the BRC API. 2. The BRC service interacts with the NetWorker server to locate the volumes that contain the requested data. 3. In certain cases, the nsrsnap_save program on the data mover computer sends the data through the BRC API to the nsrsqlrc program for recovery. This occurs in cases where the NetWorker client with access to the snapshot backup is not the target SQL Server host for the recovery, a different NetWorker client must be used as the data mover.
Restore types
To restore data, the NMSQL requires that a restore type be specified. The restore type is based on the level of backup created, as well as the set of data you need to restore from a backup. The restore type must be specified before browsing and selecting objects for the restore. The NMSQL supports four restore types: normal, partial (piecemeal) verify-only, and copy restore, depending on the version of Microsoft SQL Server software involved, as shown in Table 1-3 on page 1-15.
Table 1-3
Restore types supported with NetWorker Module for SQL user program Restore type Microsoft SQL Server version SQL 7.0 SQL 2000 SQL 2005 Enterprise ed. SQL 2005 Standard ed. Workgroup ed. Express Normal x x x x x x x x Partial Piecemeal Verify x x x x x x Copy x x x x x x
Recovery operations
1-15
Overview
Normal restore type The normal restore type restores the entire set of data associated with one or more SQL Server 7.0 or later backups, including full, incremental, and differential backups. The normal restore type supports recovery of a file, filegroup, or a database to the database originally backed up. The normal restore type can restore level full, level 1 (differential), and level incremental backups in the order required by SQL Server. When performing a restore, the NMSQL uses the normal restore type as the default. Because the NMSQL can back up only specified files and filegroups of a SQL Server 7.0 or later database, restoring from a media failure that affects only a subset of the data in the databases is possible. In addition, a single filegroup, or multiple filegroups and/or files, can be restored from a full database backup. Partial restore type The partial restore type restores a portion of the filegroups associated with a single SQL Server 2000 database backup. When a partial database restore is performed, the primary filegroup and associated files are always restored, in addition to the files specified for restore. The primary filegroup contains information necessary for restoring the database to the proper structure. Files or filegroups not selected are created, but are empty. Only a single item can be marked for this operation. In addition, a copy of a system database can be marked, but it cannot be overwritten. Piecemeal restore type A piecemeal restore, released with SQL Server 2005, is the next generation of the partial restore.
Note: The piecemeal restore is supported only with the Enterprise edition of SQL Server 2005.
Piecemeal restore is a multi-stage process that enables you to restore filegroups incrementally to a new or existing database. The first stage of a piecemeal restore includes the primary filegroup and any number of secondary filegroups, which is similar to the partial restore. After the primary filegroup is restored, the database can be brought online and additional filegroups can be restored as needed.
1-16
Overview
Verify-only restore type The verify-only restore type verifies only the backup media for the selected SQL Server 7.0 or later backups. Selecting the verify-only restore type does not restore the SQL Server data. In addition, when verify-only is specified, item-level properties for file, filegroup, and database objects are not available. Copy restore type A copy restore is an operation in which data is recovered to a SQL Server host other than the one from which it was backed up. Note that copy restore from and to the same SQL Server instance also can be done. The copy restore type creates a copy of a database by restoring a SQL Server 7.0 or later database backup to a new location, or to a new database name. The copy restore type makes it easy to replicate a database that was previously backed up. You can only mark a single item for this operation. In addition, you can copy a system database, but you cannot overwrite it. The NMSQL enables you to restore Snapshot backups to a SQL Server host that does not have a Snapshot-capable storage subsystem. Copy restore of Snapshot backups A copy restore of a Snapshot backup supports the creation of a new database for the following scenarios:

Copy restore of an instant backup to a new location. Copy restore of an instant backup to another database on the same host within the same storage array. Copy restore of an instant backup to a new database on the same host within a different storage array. Copy restore from a rollover. Copy restore to another database on the same host. Copy restore to another database on a different host.
Rollback restore type
NMSQL backups can use the PowerSnap Module functionality to create a PIT copy of a file system that belongs to a SQL Module client. This is considered an instant backup since the copy serves as the backup. Many instant backups can be performed in a single day, thus reducing the exposure to data loss.
Recovery operations
1-17
Overview
When a PIT copy is created, a unique save set ID is assigned. In addition, when the data from that PIT copy is backed up to a tape or disk medium, a different save set ID is assigned to that data. By having two unique save set IDs, the snapshots (PIT copies) can be handled separately from the backed-up data. Both are stored until either the PIT or rollover expiration policy setting. The copies of the data are removed only when all snapshots and backups of the data have been deleted. A rollback recovers a specific PIT copy to one or more volumes. You can request a rollback without having to retrieve data from a secondary storage system. Rollback of a managed or non-managed volume prevents the snapshot from being maintained and causes the snap set to become invalid. To minimize risk to data, first perform a tape backup of the snapshot before performing a rollback operation. Rollbacks are destructive by nature, which means that the entire contents of the file system is overwritten. As a default safety check, a rollback can only restore the original volume. Chapter 4, Restoring SQL Server Data provides more information.
Note: The Microsoft SQL Server module has the ability to detect volume overlap. A warning is displayed if a database is being restored to volumes that have files from other databases.
Restore modes
To restore a database, the NMSQL requires that a restore mode be specified. A restore mode instructs the SQL Server how to interact with the database after the restore operation completes. For instance, restore modes can leave the database in an intermediate state, so that additional transaction logs can be applied. Restore modes correspond to SQL Server database restore options and include: normal, no-recovery, and standby. Normal restore mode The normal restore mode instructs SQL Server to leave the database in an operational state after the restore completes. This then enables database reads and writes. The normal restore mode is the default mode the NMSQL uses when restoring a database.
1-18
Overview
No-recovery restore mode The no-recovery restore mode activates the SQL Server NORECOVERY database restore option for the last stage restored. The no-recovery restore mode places the database in an unloadable state after the restore, but is still able to process additional transaction log restore operations. Standby restore mode The standby restore mode activates the SQL Server STANDBY database restore option for the last stage restored, which forces the database to be in a read-only state between transaction log restore operations. The standby restore mode provides an undo file for SQL Server to use when rolling back the transactions. Online restore mode SQL Server 2005 provides the ability to perform a restore operation while a SQL Server database is active. The database is completely offline only while the primary filegroup is being restored. Once the primary filegroup is restored, the database can be brought online while the rest of the filegroups are being restored, and then only the data that is being restored is unavailable. The rest of the database remains available during this type of restore. Earlier versions of SQL Server require that you bring a database offline before you restore the database.
Restore time
Microsoft SQL Server and the NMSQL enable you to restore backups to a specific restore time. The restore time controls which backup data should be reinstated when a database is restored. The restore time may also control which portions of a level incremental backup are to be restored when the NMSQL is instructed to discard transactions performed after a given time. The default or current restore time for each database comes from the create time of the marked item. By default, the most recent backup is restored. If the most recent backup is level incremental or 1, dependent backups are restored first. User-specified restore times can restore older backup versions or perform point-in-time restore operations. For instance, a point-in-time restore may be specified by using a restore time that is earlier than the create time of the transaction log backup, but later than the create time of the previous backup.
Recovery operations
1-19
Overview
The NMSQL provides three methods for restoring to a specific time: database backup versions, point-in-time restore of a transaction log (level incremental) backup, and restoring to a named log mark.
NetWorker User for SQL Server program

The NetWorker User for SQL Server program is a graphical user interface that is used to run manual backups and restores of SQL Server data objects. The NetWorker program displays the data items in the SQL Server storage hierarchy for the selected SQL Server instance in the Backup and Restore windows. The Backup and Restore windows are split into two panes. The left pane consists of expandable trees that display the SQL Server storage hierarchy found on the current NMSQL host. The NetWorker User for SQL Server program enables you to browse the filegroups and files contained in a database. However, the program may display certain data items that are contained in the storage hierarchy, but are not available for backup or restore operations. These are referred to as fake objects.
Fake objects
When a file or filegroup exists in the SQL Server storage hierarchy, but cannot be backed up because of SQL Server settings on the database, the item is displayed in the Backup window by using the fake filegroup or fake file conventions outlined in Figure 1-5 on page 1-20. In addition, if you try to mark a fake object, an error dialog box appears.
Figure 1-5
1-20
Marking a fake object
Overview
Microsoft SQL Server recovery models on page 6-2 provides more information about the constraints Microsoft SQL Servers 2000 and 2005 enforce that determine whether an item is available for backup or restore.
Display conventions
The NMSQL uses specific data item names, text characteristics, and icons to distinguish the variable qualities of SQL Server data. Figure 1-4 on page 1-14 outlines these conventions.
Table 1-4
SQL Server storage hierarchy display conventions Data item SQL Server Description Root of the storage hierarchy Signifies all SQL Server databases on the host Descendant of root Signifies a database May contain filegroups Descendant of a database Signifies a database filegroup May contain files Signifies that the filegroup cannot be selected for backup Descendant of a filegroup Signifies a database file Signifies that the file cannot be selected for backup Pane Left only Icon
Database
Left and right
Filegroup
Left or right
Fake filegroup File
Left or right
Right only
Fake File
Right only
1-21
Overview
Marking items
The NMSQL provides the following methods for marking and unmarking items for backup or restore operations:
Click the checkbox to the left of the item to mark it. To unmark it, click the checkbox again. Click the item to select it, and then click the Mark button in the toolbar. To unmark it, click the Unmark button in the toolbar. Right-click the item to mark it and select Mark from the shortcut menu. To unmark it, select Unmark from the shortcut menu. Click the item to select it, and then choose Mark from the File menu. To unmark it, choose Unmark from the File menu.
Note: Some SQL Server data items are visible, but cannot be marked. For example, when certain database options are set, SQL Server does not allow file or filegroup backups. These items are displayed by using visual and behavioral cues to distinguish them from normal data items. Table 1-4 on page 1-21 provides more information on display conventions.
The NMSQL provides marking indicators that help determine the state of each item in the browse tree:
Unmarked An unmarked item is one that is not selected for backup or restore. An empty checkbox appears to the left of each unselected item to indicate it is unmarked, as shown in Figure 1-6 on page 1-22.
Figure 1-6
Unmarked items
1-22
Overview
Marked A marked item is one that is selected for backup or restore. A check mark appears in the checkbox to the left of each marked item, as shown in Figure 1-7 on page 1-23.
Figure 1-7
Marked items
Partially marked A partially marked item is one that has marked descendants, but the item itself is not explicitly marked. A partially marked item is not backed up or restored. A check mark appears in a gray check box to the left of each partially marked item, as shown.
Figure 1-8
Partially marked items
1-23
Overview
Marking semantics and restrictions

To support the browsing capabilities, the NMSQL imposes certain semantics and restrictions regarding how items may be marked. Whether an item can be marked is based on the mark status of that items predecessors and descendants. Depending upon what is marked, message dialog boxes may appear to provide additional information on the current marks and the type of operation. The NMSQL enables you to do the following:

Mark a single file, filegroup, or database. Mark multiple, heterogeneous items. Mark an item when any of that items surrounding items are already marked. Mark or unmark all SQL Server data by right-clicking the root item and selecting Mark All Databases or Unmark All Databases from the shortcut menu. Unmark all databases from the SQL Server root.
The NMSQL imposes the following restrictions:
You cannot mark an item if any of the predecessors of descendants are already marked except in the SQL Server root. When a database is marked, all of the items descendants are not automatically marked. When a filegroup is marked, all of the files are not automatically marked.
Restore window restrictions

In the Restore window, the rules for marking an item are based on the selected restore type. The normal and verify restore types do not restrict marking in any way. All restorable objects (file, filegroup, database) are markable. When the partial, piecemeal, or copy restore type is chosen, only one database object can be marked. Marking the root SQL Server item is not permitted.
When the partial or piecemeal restore type is chosen, the subset of filegroups and files of the selected database must be marked by using the Properties dialog box. Task 4: Set the restore properties (optional) on page 4-36 provides more information about the Properties dialog box.
1-24
Overview
For piecemeal restore, several of the selections you may make in the Properties dialog box will be reset if you revisit the Properties dialog box again before starting the restore process. The selections that will be reset include: Marked files and filenames of the selected database. Name for restored database option (Files tab). Back up the active portion of the transaction log before restoring the database checkbox (General tab). Redisplaying the Properties dialog box in this type of restore causes the previous selections of these options to be removed. Figure 1-9 on page 1-25 identifies the message displayed when this occurs.
Figure 1-9
Restore Options error message
When the copy restore type is chosen, filegroups and files of the selected database are automatically marked and restored as part of the full database restore.
Error logs for backup and recovery

To help you diagnose problems, the NMSQL writes the following types of information to an application-specific log file during backup and restore operations:

Software configuration information Operation parameters Operation status and error messages
The log files are written into the nsr\applogs folder on the SQL Server host. The log files are cumulative and are appended each time the nsrsqlsv or nsrsqlrc program runs. Log space management is crucial because the log file is truncated when disk space is exhausted. Table 1-5 on page 1-26 cross-references the program and log file names.
Error logs for backup and recovery
1-25
Overview
Table 1-5
Program and log file names Program nsrsqlsv nsrsqlrc XBSA library code Log file
nsrsqlsv.log nsrsqlrc.log xbsa.messages
The logging capabilities of nsrsqlsv and nsrsqlrc are not cluster-aware. For both clustered and nonclustered configurations, the logs are stored on a local disk. The xbsa.messages file collects messages from the X-Open Backup Services application programming interface.
1-26
2
Invisible Body Tag
Manual Backups
This chapter explains how to use the NetWorker User for SQL Server program to manually back up Microsoft SQL Server databases. It includes the following sections:

About manual backups .....................................................................2-2 Performing a backup with NetWorker User for SQL Server .......2-3
Read this chapter and Chapter 4, Restoring SQL Server Data before performing a backup or restore operation by using the NMSQL. The type and level of backup impacts the type of restore that can be performed. For example, regularly scheduled, full-level backups are required to enable recovery from a disaster situation.
CAUTION If you are running the NMSQL on Windows Server 2003, make sure that you have the appropriate privileges before running a backup. If you do not have administrative privileges, then run NMSQL from the console, not from Remote Desktop/terminal services. Otherwise failures will occur even if all permissions are correct. Appendix B, NetWorker Module Permissions Wizardprovides further information on setting permissions.
Manual Backups
2-1
Manual Backups
About manual backups

You can initiate a manual backup of Microsoft SQL data at any time. A manual (unscheduled) backup can be started immediately and is sometimes referred to as on-demand or ad hoc. When a manual backup is run from a NetWorker server that is a client host, only data stored on that SQL Server host can be backed up.
Note: Manual, or ad hoc backups are performed using the traditional method of backing up. You cannot start a snapshot backup manually.
The following combinations of data objects can be backed up by using the NMSQL:

The entire SQL Server storage hierarchy One or more entire databases One or more filegroups in a SQL Server 7.0 or later database One or more files in a SQL Server 7.0 or later filegroup A heterogeneous collection of files, filegroups, and databases Transaction log backups
The storage hierarchy is defined as the database storage components exposed to third-party backup vendors by the SQL Server Storage Engine. The storage components include files, filegroups, databases, and transaction logs. You can perform a manual backup by using either of the following NMSQL interfaces on the SQL Server:
The NetWorker User for SQL Server program. Performing a backup with NetWorker User for SQL Server on page 2-3 provides instructions.
Note: A manual backup started from the NetWorker User for SQL Server program can only be performed at level full.
The nsrsqlsv command from the command prompt. Using the nsrsqlsv command on page A-3 provides command syntax. To run the nsrsqlsv command, you must log into an account on the NetWorker client host that has SQL Server administrative privileges.
2-2
Manual Backups
When performing a manual level-full backup of a file or filegroup, also perform a database incremental-level backup to maintain the validity of the transaction log.
Note: For maximum data protection, perform regular, scheduled NetWorker backups of SQL Server databases. Chapter 3, Scheduled Backups provides details on running scheduled backups.
The best way to protect Microsoft SQL data is to schedule regular backups to start automatically at a set time. Manual backups are generally performed only under special circumstances, such as when you first set up the NMSQL and want to test the configuration. Due to the complexity of configuring scheduled backups by using the NetWorker Administration program, you should first either perform a traditional manual backup, or use the NetWorker Configuration Wizard to configure a basic scheduled backup, before proceeding to configure scheduled backups.
Performing a backup with NetWorker User for SQL Server

The NetWorker User for SQL Server program is a graphical user interface that is used to perform manual backups and restores initiated by the client. The Backup window of the NetWorker User for SQL Server program displays data that is available for backup based on the SQL Server database settings. SQL database items that cannot be backed up are not displayed in the Backup window. This includes, but is not limited to, databases in the following states:

Standby Offline Not recovered Loading Prerecovery Single user with active user connections
Note: The NMSQL does not support manual snapshot backup from either NetWorker User for SQL Server program or Command prompt.
2-3
Manual Backups
To perform a manual backup, complete the following tasks:
Task 1: Start the NetWorker User for SQL Server program on page 2-4 Task 2: Select the SQL Server data on page 2-6 Task 3: Set the backup options on page 2-7 Task 4: Set the backup properties for each marked database on page 2-10 Task 5: Start and monitor the backup on page 2-11 Task 6: Back up the client indexes and bootstrap file on page 2-12
Task 1: Start the NetWorker User for SQL Server program

To start the NetWorker User for SQL Server program: 1. From the Start menu, select Programs > NetWorker > NetWorker User for SQL Server. If multiple instances of SQL Server are active on the computer, the Select SQL Instance dialog box opens before the main window opens.
Figure 2-1
Select SQL Instance dialog box
2-4
Manual Backups
2. Select the SQL Server instance that the NetWorker Module will back up, and click OK. NetWorker User for SQL Server connects to the selected instance. The main window appears.
Figure 2-2
NetWorker User for SQL Server main window
3. (Optional) To select a NetWorker server other than the default server that was specified during the NMSQL installation. a. Click the Select NetWorker Server button on the toolbar.
2-5
Manual Backups
The Change Server dialog box appears.
Figure 2-3
Change Server dialog box
b. Select a NetWorker server from the list, and click OK.
Task 2: Select the SQL Server data

When performing a backup by using the NetWorker User for SQL Server program, the NetWorker server always performs a full backup of the SQL Server data. To select the SQL Server data to back up from the NetWorker User for SQL Server program: 1. Select Backup from the Operation menu.
2-6
Manual Backups
The Backup window appears and displays a hierarchical list of SQL Server data objects available for backup.
Figure 2-4
Backup window
2. To expand an object, click the plus sign (+) beside the object name in the left pane. The descendants of the object appear in the right pane. 3. Mark one or more objects for backup. Marking items on page 1-22 provides complete details on marking or unmarking items. Some SQL Server data items are visible in the Backup window, but cannot be marked. If certain database options are set, SQL Server prohibits a file or filegroup backup. These unavailable objects appear dimmed to distinguish them from those that are available to back up. Display conventions on page 1-21 provides more information on these icons.
Task 3: Set the backup options

To set backup options: 1. Select the required attributes in the Backup Options dialog box before starting the backup. The Backup Options dialog box appears as shown in Figure 2-5 on page 2-8
2-7
Manual Backups
The selected attributes apply to each marked object for all backup operations performed until the NetWorker User for SQL Server program is closed. Selected pools, however, are stored in the registry and persist from one session to another. If an attribute is not selected, then the NetWorker User for SQL Server default is used.
Figure 2-5
Backup Options dialog box
2. On the General tab, select the appropriate attributes: Compress the backup content Applies XBSA compression to all marked databases before writing the backup data to the storage device. In the same manual backup, certain databases cannot be backed up with compression and others without. Compressing data for a backup generates less network traffic and uses less backup media space, but it consumes additional CPU resources. Most tape devices perform compression, which makes software compression unnecessary.
2-8
Manual Backups
Create a striped backup Creates a striped backup by using the SQL Striped feature. If this attribute is checked, the Stripes list is enabled. Appendix C, Striped Backup and Recovery provides more information about striping. To select the number of striped for backup, select a number from the list box. The maximum number of stripes the NetWorker Module software supports (32) is the same for SQL Server 7.0, 2000, and 2005. However, the maximum number of stripes cannot be more than the value set for NetWorker parallelism. Use pools for media management Lists media volume pools for storing the backup contents. This attribute applies only to manual backup operations performed from the NetWorker User for SQL Server program or from a command prompt. If this attribute is checked, the Full Backup Pool and the Log File Pool lists are enabled. To select volume pools: a. Check the Use Pools for Media Management attribute. b. Select a pool from the Full Backup Pool list to store full SQL Server backups, including databases and filegroups. c. Select a pool from the Log File Pool list to store transaction log or differential (level 1) SQL Server backups.
Note: The pool names in the Full Backup Pool or Log File Pool lists are initially created by using the NetWorker Administrator program, and are stored in the operating system registry.
Select debug level Defines the level of debug information to be sent to the backup status window during the backup operation. Levels range from 1-9, with 1 representing the least amount of information. 3. Click OK to close the Backup Options window.
2-9
Manual Backups
Task 4: Set the backup properties for each marked database

Set backup properties for each marked database before starting the backup operation. If a property is not specified, the default is used. After the backup operation is complete, and the Backup window is closed, the property values revert back to the NetWorker User for SQL Server defaults. To set the backup properties for each marked database: 1. Mark one or more databases in the Backup window, and then right-click each marked database and select Properties from the shortcut menu. The Properties dialog box appears.
Figure 2-6
Properties dialog box
2-10
Manual Backups
2. Select the appropriate options: Truncate content before performing backup Truncates the transaction logs before backup. Perform checksum before writing to media Performs a checksum operation with the backup and saves the information to the backup media. Another checksum is performed before a restore to ensure that it matches the backup. A checksum is used to detect a partial backup or restore state. The NMSQL verifies the checksum by calculating a local result and comparing it with the stored value. If the values do not match, you can choose to continue the backup or restore operation. Select the Continue on checksum error option to backup and restore if errors are detected.
Note: This option is available with SQL Server 2005 instances.
3. Click OK.
Task 5: Start and monitor the backup

A backup cannot run if there is no media volume mounted in the backup device. Before starting a backup, ensure that a labeled media volume is mounted in the backup device. The volume should be labeled for the volume pool where the backup is to be directed. If there is no volume in the backup device when a backup is started, no messages appear in the Backup Status window and the backup waits for operator intervention. To start the backup: 1. Select Start Backup from the File menu. 2. Monitor the backup messages in the Backup Status window. After the backup is finished, a Backup Completed message appears.
Note: The amount of time to back up a database depends on database size, network traffic, server load, and tape positioning.
2-11
Manual Backups
3. Close the Backup Status window. To cancel a backup, select End Backup from the File menu.
Task 6: Back up the client indexes and bootstrap file

Performing a manual backup of SQL Server data does not automatically back up the client indexes and bootstrap file. To back up client indexes and a bootstrap file: 1. Log in as one of the following: As root on a UNIX NetWorker server. As administrator on a Windows NetWorker server. 2. Enter the following command from the command line:
savegrp -O -l full -P printer_name -c NetWorker_client
where: printer_name is the name of the printer where the bootstrap information is printed at the end of the bootstrap backup. NetWorker_client is the hostname of the SQL Server.
2-12
3
Invisible Body Tag
Scheduled Backups
This chapter explains how to configure scheduled backups of SQL Server data. This chapter includes the following sections:

About scheduled backups ................................................................3-2 Setting backup levels .........................................................................3-3 Configuring scheduled backups ......................................................3-7 Unavailable databases and group backup failure.......................3-19
CAUTION If you are running the NMSQL on Windows Server 2003, make sure that you have the appropriate privileges before running a backup. If you do not have administrative privileges, then run the NMSQL from the console, not from Remote Desktop/terminal services. Otherwise failures will occur even if all permissions are correct. Appendix B, NetWorker Module Permissions Wizard provides further information on setting permissions.
Scheduled Backups
3-1
Scheduled Backups
About scheduled backups

The most reliable way to protect Microsoft SQL data is to schedule backups of the SQL Server to run at regular intervals. Scheduled backups ensure that all SQL Server data is automatically saved, including the NetWorker servers client indexes and bootstrap file. The client indexes and bootstrap file are vital for restoring data to the SQL Server in the event of a disaster. Scheduling backups for the NMSQL is similar to scheduling NetWorker file system backups. On the NetWorker server, appropriate attribute values must be set for various resources, such as Policy, Group, and Client resources. This is done by an administrator with a working knowledge of the NetWorker software through the NetWorker Administrator program. Scheduled backups can be configured to run at any time and use backup levels full, incremental, and differential (level 1-9). An alternative method of scheduling backups is to use the NetWorker Configuration Wizard. NetWorker release 7.3 includes the configuration wizard. For versions of NetWorker release 7.2, the wizard is available as an add-on tool. The configuration wizard enables you to create one or more client or device resources for basic scheduled backups of NetWorker release 7.2 and later client computers and Microsoft SQL Server data. The wizard integrates with the NMSQL by prompting users for information, such as:

Backup type Objects to back up Recovery level Schedule preferences Microsoft SQL administrator account information
Once the wizard creates a resource, that resource can then be edited by using the NetWorker Administrator program. To use the NetWorker Configuration Wizard to configure scheduled backups, first install NetWorker release 7.2 or later client software, and then separately install the NetWorker Configuration Wizard. The NetWorker Installation Guide and NetWorker Release Notes provides instructions on how to install and use the NetWorker Configuration Wizard.
3-2
Scheduled Backups
Note: The wizard functionality is available by default in NetWorker release 7.3, and need not be installed separately for NetWorker release 7.3.
Once the configuration wizard is installed, with the NetWorker 7.3 software, you can then install the NMSQL. During installation of the NMSQL software, you can choose to install the NMSQL for Microsoft SQL Server component from the wizard. The NetWorker Module for Microsoft SQL Server Installation Guide provides installation instructions.
Setting backup levels

The NMSQL enables you to specify these backup levels, along with database full, database differential, and database incremental. The availability of a backup level depends on the type of data selected for backup and any SQL Server settings on those objects. Figure 3-1 on page 3-14 identifies the supported backup levels.
Table 3-1
Backup levels for SQL Server data SQL Server 7.0 SQL Server 2000, 2005 incr yesc, yesa,c n/a n/a n/a full yes yes yesa,c yesa,c yesa,c diff yesc yesc yesa,c yesa,c yesa,c incr yesc yesa,c n/a n/a n/a
SQL Server data objects All databases, including MSSQL: Specified databases All filegroups in specified databases Specified filegroups in specified database Specified files in filegroups in specified databases
full yes yes yesa,c yesa,c yesa,c
diff yes yesc noa,b,c noa,b,c noa,b,c
a. An incremental file, filegroup, or database backup can only be created when the SQL Server database options are properly configured. For more information, refer to the Microsoft SQL Server 7.0, 2000, or 2005 documentation. b. SQL Server 7.0 does not allow differential (level 1) backups of filegroups or files. If a differential file or filegroup backup is specified, the NMSQL performs a full backup. c. Individual items are subject to promotion. Promoting backup levels on page 3-6 provides more information.
3-3
Scheduled Backups
Strategies for backing up SQL Server data If the SQL Server manages a significant amount of data, schedule a backup of the databases every one to two weeks, as shown in Table 3-2 on page 3-4.
Table 3-2
Full backup every one to two weeks Fri full incr full Sat incr incr Repeat Sun incr incr Mon incr diff Tues incr incr Wed diff incr Thurs incr incr
Another backup strategy is to schedule incremental backups on several successive days immediately following the previous full backup, as shown in the following table. This schedule backs up all data that has changed since the previous incremental backup.
Table 3-3
Incremental backup after a full backup Fri full Repeat Sat incr Sun incr Mon incr Tues diff Wed incr Thurs incr
A level 1 differential backup can also be scheduled after several days of incremental backups. This schedule backs up all data since the previous full backup.
Note: If a database has been made read-only, a full backup of the database should be made. A read-only database cannot be restored from a transaction log backup that may already exist.
NetWorker Administrator Guide provides further details and examples of planning backup strategies and creating schedules.
Differences between backup levels

Because it may not be practical or efficient to run full backups every day, other backup levels can be specified for automatic, scheduled backups. Limiting the frequency of full backups can decrease server load while ensuring data is protected.
3-4
Scheduled Backups
Consider the following when selecting backup levels:
Full backups take more time than differential backups, while differential backups take more time than incremental backups. However, restoring data immediately following a differential backup is generally faster than restoring data following a number of successive incremental backups. If you have only a stand-alone storage device and a full backup does not fit on a single media volume, an operator must monitor the backup. This way, the volume can be changed at the appropriate time. An incremental backup saves only transactions that have occurred since the most recent full or transaction log backup. For this reason, using incremental backups can simplify and expedite database recovery.
Table 3-4 on page 3-5 outlines the differences between the backup levels:
Table 3-4
Backup level advantages and disadvantages Backup level Full Advantages Fastest restore time. Disadvantages Slow backup. Increases load on client, server, and network. Uses the most volume space. Slow restore. Data can spread across multiple volumes. Multiple transaction logs can spread across multiple volumes. Generally more time-consuming than a incremental backup (depending on the backup schedule strategy).
Incremental (transaction log)
Faster than a full backup. Decreases the load on server and Uses the least volume space. Enables point-in-time restore. Faster than a full backup. Captures all changes since last full.
Differential
3-5
Scheduled Backups
Combining data objects to create backup levels The NMSQL enables you to select SQL Server data objects in various combinations to create scheduled backups of different levels, as shown in Table 3-5 on page 3-6.
Table 3-5
Creating additional backup levels with data objects Backup level Full database Full file or filegroup Database objects Select one or more databases to create a level full database backup of the selected databases and their transaction log files. Select one or more files or one or more filegroups to create a level full file or filegroup backup of the selected files or filegroup, but not their transaction logs. Select one or more databases to create a database incremental level backup of only the incremental for the selected databases. The SQL database must be preconfigured to enable incremental backups. Database differential (level 1) File or filegroup differential Select one or more databases to create a database level differential backup of only the changes made to the selected databases since the last full-level backup was created. For SQL Server 2000 and 2005 only, select one or more files, or one or more filegroups to create a file or filegroup level differential backup of only the changes made to the selected files or filegroups since the last full level backup was created.
Database incremental
Promoting backup levels Guidelines for Microsoft SQL Server Best Practices indicate that a full database backup should be the first step in implementing a recovery strategy for a database. In adhering to these guidelines, the NetWorker Module supports backup level promotion. Table 3-6 on page 3-6 explains what prompts a promotion.
Table 3-6
Backup level promotion process Item Database Requested level Differential Level of promoted Full Reason Database full backup does not exist.
3-6
Scheduled Backups
Table 3-6
Backup level promotion process (continued) Item Database Requested level Incremental Level of promoted Full Reason Database full backup does not exist. Database does not support incremental (transaction log) backups.a Database is currently in emergency mode.a Full backup of the entire database does not exist.b Full backup of the entire database does not exist. SQL Server 7.0 does not support file/filegroup differential backups. File/filegroup incremental backups are not supported.
File/Filegroup File/Filegroup File/Filegroup when connected to SQL Server 7.0 File/Filegroup
Full Differential Differential
Database full Database full Full
Incremental
Full
a. Refer to the Microsoft SQL Server Books Online for more information. b. Databases consist of files and groups that contain files. The default configuration is a primary filegroup with the main data file. Elaborate database configurations can contain more filegroups; each with more files. If a filegroup or file level backup is specified, and a full database backup is not on record, the filegroup or file backup is promoted to a database full backup.
Configuring scheduled backups

Configure scheduled backups by using either the NetWorker Configuration Wizard or the NetWorker Administrator program. For instructions on using the wizard to configure a basic scheduled backup, refer to the NetWorker Release Notes. To configure scheduled backups by using the NetWorker Administrator program, complete the following tasks:

Task 1: Assign a Snapshot Policy (optional) on page 3-8 Task 2: Configure one or more group resources on page 3-10 Task 3: Configure one or more Client resources on page 3-11
3-7
Scheduled Backups
Task 4: Configure a schedule and set backup levels on page 3-13 Task 5: Configure the NetWorker volume pools on page 3-17 Task 6: Test the Configuration on page 3-18
Before beginning these tasks, make sure the NetWorker Administrator program is configured to display hidden attributes. To display hidden attributes:
For NetWorker 7.2 and earlier servers on Windows platforms: Select Customize from the Options menu. On the Window Configurations tab of the Customize dialog box, select Display Hidden Attributes.
For NetWorker 7.2 and earlier servers on UNIX or Linux platforms, select Details from the View menu. For NetWorker 7.3 and later servers, hidden attributes are called diagnostic attributes. To display diagnostic attributes in the Administration window, select Diagnostic Mode from the View menu.
The NMSQL can back up to a NetWorker server that is running on any supported operating system. The appropriate version of NetWorker Administrator Guide provides instructions on using the NetWorker Administrator program to configure NetWorker server resources.
Task 1: Assign a Snapshot Policy (optional)

If the NetWorker PowerSnap Module is installed, assign a preconfigured or custom snapshot policy on the NetWorker server. This guide provides only the snapshot configuration information that is relevant to the NMSQL.
The NetWorker Administrator Guide provides information about creating a custom snapshot policy. The NetWorker PowerSnap Module Installation and Administrators Guide appropriate for the storage subsystem provides instructions on setting the required NetWorker server resources for snapshot operations.
3-8
Scheduled Backups
Using NetWorker 7.2 and earlier To assign a snapshot policy on the NetWorker server: 1. In the NetWorker Administrator program, click Manage Groups from the Configure tab. 2. Right-click the icon of the appropriate group, and select Edit. 3. Click the Miscellaneous tab and select a preconfigured or custom snapshot policy. Table 3-7 on page 3-9 shows a sample snapshot policy that creates four snapshots per day. Only the first snapshot is moved to secondary storage. All snapshots are deleted from primary storage after 24 hours.
Table 3-7
Sample snapshot policy Name Comment Number of Snapshots Snapshot Expiration Policy Retain Snapshots Backup Snapshots Four Snaps Daily Snapshot policy for SQL Server backups 4 4 Day First
Using NetWorker 7.3 and later To create a snapshot policy: 1. In the Administration window, click Configuration. 2. In the expanded left pane, select Snapshot Policies. 3. From the File menu, select New. 4. In the Create Snapshot Policy dialog box, enter a name for the snapshot policy in the Name attribute and complete other attributes as appropriate.
Note: NetWorker PowerSnap Module documentation provides information about how to complete the attributes for a snapshot policy.
3-9
Scheduled Backups
5. Click OK. To assign a snapshot policy to a backup group: 1. In the Administration window, click Configuration. 2. In the expanded left pane, expand Groups, right-click the appropriate group, and select Properties. 3. In the Setup tab, select the appropriate snapshot policy.
Task 2: Configure one or more group resources

A NetWorker backup group is a set of NetWorker Client resources, all of which start backing up data at a specified time once the Autostart feature is enabled and the backup start time is specified in the Group resource. You can assign one or more SQL Server hosts to a NetWorker backup group. NetWorker software provides a preconfigured group named Default. The Default groups attributes can be modified, but the Default group cannot be deleted from the list of NetWorker groups. The Default group has the following attributes:

Autostart = Disabled Start time = 3:33 Client retries = 1 Clones = No Clone pool = Default Clone Interval = 24:00 Snapshot = False Snapshot Policy = Daily Snapshot Pool = Default
To assign the SQL Server host to another group with different attributes, create a group in the NetWorker Administrator program before creating a NetWorker Client resource for the SQL Server host. The NetWorker Administrator Guide provides complete instructions on creating backup groups.
3-10
Scheduled Backups
To back up several large SQL Server databases, consider creating a separate backup group with a different start time for each database. Any number of backup groups can be set up in the NetWorker Administrator program. Use of backup groups can help reduce network traffic or load on the NetWorker server. When selecting a start time for each group, schedule the groups far enough apart for one group to complete its backup before the next group starts. Choose times when there is little network activity, for example, nights and weekends.
Task 3: Configure one or more Client resources

A NetWorker Client is a resource configured on the NetWorker server. This resource defines the following information:

Client data to back up Backup schedule for the client Browse policy for the backup data Retention policy for the backup data
Each SQL Server host to be backed up must be configured as a NetWorker client in the NetWorker Administrator program. In addition, multiple SQL Server databases that exist on the same SQL Server host can be configured as separate NetWorker clients. For each Client resource, the NetWorker server does the following:
Maintains the Client resource information, including entries in the online client file index and media database. Contacts the clients listed in a backup group configured on the server. Performs the scheduled backups when a client request is received. Restores the data upon request from the client.
3-11
Scheduled Backups
Table 3-8 on page 3-12 describes the settings used when creating a NetWorker client for a SQL Server host.
Table 3-8
Client resource attribute settings Attribute Name Comment Save Set NMSQL Requirements Enter the SQL Servers hostname. If you create multiple Client resources for the same SQL Server, use the same name for each. If using multiple Client resources for the same SQL Server host, enter a comment to identify the purpose of each. Specify any valid save set names, for example: MSSQL: or MSSQL:dbName1 [MSSQL:dbName2 MSSQL:dbName3 ...] For example, entering only MSSQL: always yields a backup of all databases on the SQL Server host. A snapshot backup fails if more than one database, or MSSQL: is specified for the Save Set attribute. For snapshot backups, list only one database for the Save Set attribute. Select a backup group. Task 2: Configure one or more group resources on page 3-10 provides more information. Select a backup schedule. Task 4: Configure a schedule and set backup levels on page 3-13 provides more information. Select a browse policy to specify how long the NetWorker server retains client file index entries. Select a retention policy to specify how long the NetWorker server retains media entries for the clients backups. If the NetWorker server has one or more remote storage nodes that are to be used for backing up the SQL Servers data, enter the name of each storage node in the order they are to be used. The default storage node, nsrserverhost, represents the NetWorker server. Enter the nsrsqlsv command and any necessary command options. Command options for nsrsqlsv on page A-5 provides information about nsrsqlsv options. For virtual server backups, the -a virtual_server_name command option is required.
Group Schedule Browse Policy Retention Policy Storage Node
Backup Command
3-12
Scheduled Backups
Table 3-8
Client resource attribute settings (continued) Attribute Remote Access NMSQL Requirements Enter the user ID or hostnames of other clients for the Remote Access attribute. This grants to those hosts copy restore type permission, which enables the named hosts to access the NetWorker server and receive directed recover data. If this attribute is left empty, only administrators and users logged on to the SQL Server host have access. For a serverless backup, this attribute must include the proxy client hostname. To enable the NMSQL to back up the SQL Server virtual server and/or a mirrored server, enter the username for a Windows user account that has SQL Server administrator privileges. For mirroring, this should be the same user account and password that were used to set up the mirroring relationship. If multiple accounts have been set up, only one needs to be specified. Per Microsoft SQL documentation, a domain account must be used to set up the mirroring relationship. Enter the password for the Remote User account. Enter all known aliases for the SQL Server host in the Aliases attribute on the Preference tab, as in the following example: mars mars.legato.com Include both the DNS short name and long name for the SQL Server host.
Remote User
Password Aliases
Task 4: Configure a schedule and set backup levels

As a NetWorker client, a SQL Servers backup schedule is controlled by a Schedule resource on the NetWorker server. The schedule determines what level of backup (for example, full or incremental) is performed on a given day. Setting backup levels on page 3-3 provides implementation details and examples. The NetWorker server provides several preconfigured schedules, such as the Default schedule shown in Figure 3-1 on page 3-14. You can use the preconfigured schedules as they are, or modify them to meet your needs. You can also create custom schedules.
3-13
Scheduled Backups
Using NetWorker 7.2 and earlier To configure a schedule: 1. From the NetWorker Administrator program, select Manage Schedules. The Schedules dialog box appears.
Figure 3-1
Select Schedule dialog box
2. Perform one of the following: To create a new schedule, right-click Schedules and select Create. To modify an existing schedule, right-click the schedule and select Edit.
3-14
Scheduled Backups
The Edit or Create Schedule dialog box appears.
Figure 3-2
Edit Schedule dialog box
3. From the Edit Schedule dialog box, select the Set Level button to set the backup level for each day in the schedule.
Figure 3-3
Set Level dialog box
The backup level specified in this dialog box takes effect when the scheduled backup runs. Setting backup levels on page 3-3 provides implementation details and examples.
3-15
Scheduled Backups
4. Assign a backup schedule to a client or a group: To assign a schedule to a group, select the schedule name in the Schedule attribute of the Group resource. Task 2: Configure one or more group resources on page 3-10 provides more information. To assign a schedule to an individual client, select the schedule name in the Schedule attribute of the Client resource. Task 3: Configure one or more Client resources on page 3-11 provides more information on the Client resource.
Note: If there is a group to be backed up at the same level every time, do not assign a schedule. Instead, select the appropriate level in the Level attribute of the Group resource.
Using NetWorker 7.3 and later To create a backup schedule: 1. From the Administration window, click Configuration. 2. In the expanded left pane, select Schedules. 3. From the File menu, select New. 4. In the Name attribute, enter a name for the schedule. 5. From the Period attribute, select Week or Month. Select Week to create a weekly backup schedule. For example, if a full backup for a Friday is selected, every Friday will have a full backup. Select Month to create a monthly schedule. For example, if a full backup for the first of the month is selected, every month will have a full backup on the first of the month. 6. Select a backup level for each day in the weekly or monthly period. a. Select a day. b. Right-click and from the Set Level menu, select a backup level.
3-16
Scheduled Backups
7. If required, select an override backup level for any day. An override occurs once only for the selected day. a. Select a day. b. Right-click and from the Override Level menu, select a backup level. 8. Click OK. To assign a schedule to a group: 1. In the expanded left pane, right-click the group name under Groups, and select Properties. 2. In the Advanced tab of the Properties dialog box, select a schedule. To assign a schedule to a client: 1. In the expanded left pane, select Clients. 2. In the right side pane, right-click a client name and select Properties. 3. In the General tab of the Properties dialog box, select a schedule.
Task 5: Configure the NetWorker volume pools

With the NetWorker server software, backups can be directed to groups of media or backup volumes called pools. A pool is a specific collection of volumes to which the NetWorker server writes data. The NetWorker server uses pools of volumes to sort and store data. The configuration settings for each pool act as filters that tell the server which volumes should receive specific data. The NetWorker server uses pools in conjunction with label templates to keep track of what data is on each specific volume. For scheduled backups only, the NMSQL uses the criteria assigned in the NetWorker Administrator program. However, pool settings that are specified in the Backup Options dialog box of the NetWorker User for SQL Server program apply only to manual backups. Task 3: Set the backup options on page 2-7 provides more information on pools used by the NetWorker User for SQL Server program.
3-17
Scheduled Backups
Each NetWorker media volume belongs either to the preconfigured Default pool or to a custom created pool. If data is not assigned to a specific pool of volumes, NetWorker software automatically uses the Default pool and searches the backup device for volumes with the pool label Default. To create, modify, or remove a Pool resource, use the NetWorker Administrator program. The NetWorker online help and the NetWorker Administrator Guide provides more information on volume pools and how to configure NetWorker Pool and Label Template resources. Task 6: Test the Configuration You can test the backup configuration by starting a backup group manually from the NetWorker Administrator program. The NetWorker server immediately backs up the clients in the group, overriding the scheduled backup start time. Each client in the group is backed up at the level defined by the schedule that is selected in the clients Schedule attribute. To write the results of a scheduled backup to a log file, enter the following in the Action attribute of the NetWorker Notification resource:
For UNIX, enter:

/usr/ucb/logger
For Windows, enter:

nsrlog -f filename
where filename is the name of a file to which the backup results are written. The NetWorker Administrators Guide provides instructions on configuring Notification resources. Using NetWorker 7.2 and earlier There are two ways to start a group manually, depending on whether NetWorker is running on Windows or UNIX. Starting a group on Windows To start a group manually on Windows: 1. In the NetWorker Administrator program, click Manage Groups on the Configure tab. 2. Right-click the group to be started, and select Start.
3-18
Scheduled Backups
3. Click Yes to start the group. Starting a group on UNIX To start a group manually on UNIX: 1. In the NetWorker Administrator program, select Group Control from the Server menu to open the Group Control window. 2. Select the name of the group to be started. 3. Click Start, then click OK. Using NetWorker 7.3 and later You can override the scheduled backup start time and start the group manually. This is equivalent to selecting Start Now in the Autostart attribute of the Group resource.
Note: When a group backup is started manually, the NetWorker server runs the backup at the level of the next scheduled backup, such as full, level [1 9], incremental, or consolidated.
To manually start a group backup: 1. From the Administration window, click Monitoring. 2. Click the Groups tab. 3. Right-click the group to start, then select Start. 4. Click Yes to confirm the start. The NetWorker server immediately backs up the clients in the group, overriding the scheduled backup start time. The group icon changes to the clock icon until the backup has completed or is interrupted.
Unavailable databases and group backup failure

The group containing a scheduled backup of a NetWorker SQL Server by using the MSSQL: saveset to back up all databases fails if any database is unavailable. This does not mean complete failure, but rather that one or more databases were not backed up successfully. A database that is in any of the following states will cause a scheduled backup to fail:

Standby Offline
Unavailable databases and group backup failure
3-19
Scheduled Backups
Not recovered Loading Prerecovery Single user with active user connections
Note: For nonscheduled manual backups that are initiated from the SQL Module on the client computer, unavailable databases are silently skipped.
Limitations of the savegrp program reporting and savegrp log file may make failure identification and the specific unavailable databases difficult to isolate. The savegrp information is displayed alphabetically, leaving some early information suppressed, and the success/failure information combined. Definitive results are available in the daemon.log file, located on the NetWorker server and in the nsrsqlsv.log file on the client computer. After the completion of a backup, the following types of error messages are listed in the daemon.log file:
Database 'Acme' cannot be opened because it is offline. Processing Acme failed, the item will be skipped. Database 'Acme' is in warm-standby state (set by executing RESTORE WITH STANDBY) and cannot be backed up until the entire load sequence is completed. Processing Acme failed, the item will be skipped.
3-20
4
Invisible Body Tag
This chapter describes the steps required to restore Microsoft SQL Server data by using the NetWorker User for SQL Server program. For guidelines on using the NetWorker Module for Microsoft SQL (NMSQL) nsrsqlrc command and command-line options to perform a restore from a command prompt, see Using the nsrsqlrc command on page A-8. This chapter includes the following sections:

Before the restore................................................................................4-2 Backing up the active portion of the transaction log ....................4-4 NetWorker Module database restore process ................................4-7 Performing a restore with NetWorker User for SQL Server ........4-8
CAUTION If you are running the NMSQL on Windows Server 2003, make sure that you have the appropriate privileges before running a backup. If you do not have administrative privileges, then you must run the NMSQL from the console, not from Remote Desktop/terminal services. Otherwise, failures will occur even if all permissions are correct. Appendix B, NetWorker Module Permissions Wizard, provides further information on setting permissions.
4-1
Before the restore

The Restore window contains database objects available for restoring. Based on the restore type selected, restore windows restrict the marking of database objects. The normal and verify-only restore types do not restrict the marking of database objects in the browse window. However, the partial (piecemeal) and copy restore types allow only database objects to be marked for restore. Restore window restrictions on page 1-24 provides more information on restore restrictions.
Note: To restore SQL Server data, use the NetWorker User for SQL Server program or the nsrsqlrc command. You cannot restore SQL Server data by using the NetWorker Administrator program, nor can you restore data that was backed up by third-party vendors.
Note: Read the Microsoft SQL Server documentation, in order to understand the limitations associated with the various recover types on the various SQL Server versions.
Before starting a restore, complete the following preparations:
Ensure that the NetWorker server software is running on the appropriate host and the NetWorker Remote Exec Service is started on the SQL Server host. If a backup of another database is in progress, wait for it to finish. Microsoft SQL Server will not restore a database while another database backup is in progress. If a SQL Server startup is in progress, wait for it to finish before starting a restore operation. Review the ERRORLOG file to determine if a database is currently being recovered or search the ERRORLOG file for the Recovery complete string. If the nsrsqlrc program is started while the SQL Server is recovering databases, the following error message appears:
Could not find database ID. Database may not be activated yet or may be in transition.
Ensure that all database users are logged off the database. A restore fails if other users try to use the database during the restore operation.
4-2
Note: In SQL Server 2005, if the primary filegroup is not under restore, then the online (piecemeal) restore functionality allows user access to a database while backup or restore is in progress, provided that the primary filegroup is not under restore.
If performing an instant restore from a snapshot-capable storage subsystem that implements copy-on-write (COW) technology, ensure that adequate space exists in the repository or cache. If the repository or cache becomes full during an instant restore, the restore fails, which causes the source SQL database to be left in an unknown state and the snapshot is deleted from the storage subsystem. The NetWorker PowerSnap Module Installation and Administrator's Guide provides information about how to ensure that the repository or cache has enough space for an instant restore operation.
Rollback restore requirements

The following should be considered before a rollback recovery can be performed:
The database files that are being recovered are the only file system objects on the volume. Additionally there should not be any other database files belonging to other databases on the volume. If there are other file system objects in the volume before the rollback, but they were not backed up by using PowerSnap, they will be damaged by the recover operation. The rollback operation checks for additional file system objects and does not start the rollback unless the force option (-F) is used, or the file system object is specified in the /nsr/res/psrollback.res file.
Note: The NetWorker PowerSnap Module Installation and Administrators Guide provides information about the psrollback.res file and force option.
Before the restore
4-3
When performing a rollback recovery, consider how this affects future snapshots. Delete the original snapshot, and all subsequent snapshots that existed prior to when the rollback was performed. This information is maintained in the client file index on the NetWorker server. If this information is not deleted the following occurs:

Future snapshots accumulate on invalid data. SQL Module restore operations fail if based on old client file index information.
The NetWorker PowerSnap Module Installation and Administrators Guide provides more information about rollback recovery.
Piecemeal restore requirements

Microsoft SQL Server 2005 Enterprise Edition databases consisting of multiple filegroups can be restored in stages with piecemeal restore. NMSQL 5.0 for Microsoft SQL Server supports piecemeal restore only with SQL Server 2005 Enterprise Edition. Partial restore in SQL Server 2000 is a one-stage process that restores part of a database to a different location. With Partial restore, the primary filegroup must be part of each partial restore. The database remains offline during the restore process. NMSQL 5.0 for Microsoft SQL Server supports partial restore for SQL Server 2000. Piecemeal restore is a multistage process that restores a database to itself or to another location. The initial stage must include the primary filegroup and optionally other filegroups. Once the primary filegroup is restored, you can bring the database online and continue restoring the remaining filegroups in subsequent stages.
Backing up the active portion of the transaction log

SQL Server 2005 requires users to perform a backup of the active portion of the transaction log prior to restore. If you are restoring a SQL Server 2005 database, first backup the active portion of the log without the recovery option. The NetWorker User for SQL Server program will automatically back up the active portion of the transaction log prior to restoring an SQL Server 2005 database.
4-4
If you are restoring a file or filegroup of a database which resides on either SQL Server 2000 instance or on SQL Server 2005 non-Enterprise Edition instance, first back up the active portion of the transaction log. The transaction log back up must be applied to the file or filegroup restore to ensure the file or filegroup is consistent with the rest of the database. If a file or filegroup is restored by using the NetWorker User for SQL Server program, this transaction log backup occurs automatically. If you are restoring a secondary filegroup (or a file belonging to a secondary filegroup) of a database residing on SQL Server 2005 Enterprise Edition, you do not need to back up the active portion of the transaction log before restoring the file or filegroup. Instead, a backup of the active portion of transaction log should be taken after restoring the file or filegroup. The transaction log backup taken should then be applied to ensure that the file or filegroup is consistent with the rest of the database. If the secondary filegroup (or file belonging to the secondary filegroup) is restored by using NetWorker User for SQL Server program, the transaction log backup occurs automatically.
Note: If you are performing a normal or piecemeal restore of a SQL Server 2005 instance, the active portion of the transaction log backup (referred to as a log tail backup) happens automatically through the NetWorker User for SQL Server program as part of the restore process. If you are performing a restore operation from the command line, you must issue two commands: The first is to back up the active portion of the transaction log. The second is to restore data.
Backup the transaction log

To use the command line to restore the file or filegroup, the transaction log backup must be explicitly performed by using the nsrsqlsv command as follows:
nsrsqlsv [-s NetWorker_server_name] -l incr -R -H dbName
where:

NetWorker_server_name is the hostname of the NetWorker server. dbName is the name of the database that receives the transaction log backup. The -R flag is required so the transaction log is not truncated after the backup completes.
Backing up the active portion of the transaction log
4-5
The -H option uses the NORECOVERY option when backing up transaction logs. It leaves the database in Restoring state. Do not use the -H option with versions previous to SQL Server 2005. Use the -H option: To back up the transaction log prior to database restore or primary filegroup restore on SQL Server 2005 Enterprise Edition. To back up the transaction log prior to database restore or filegroup or file restore on non-Enterprise editions of SQL Server 2005.
Note: If you use third-party vendor software to back up SQL Server data, after completing the backup, you must perform a full database backup with the NMSQL. This prevents a broken transaction log chain, which can cause the restore operation to fail.
Backing up the transaction log for SQL Server 7.0

Restoring a file or filegroup for a Microsoft SQL Server 7.0 database might require an additional transaction log backup. Performing the first transaction log backup for Microsoft SQL Server 7.0 could result in additional transactions on the database. If, after performing the first transaction log backup, the nsrsqlrc command fails, use the following procedure to restore the file or filegroup: 1. Put the database into single-user mode. 2. Immediately perform another transaction log backup. 3. Restore the file or filegroup. 4. Take the database out of single-user mode. The Microsoft SQL Server 7.0 documentation provides more information about file or filegroup backup and restore.
4-6
Backing up the transaction log for SQL Server Express

For SQL Server Express Edition, the recovery model of the database is set to "SIMPLE" by default, and the transaction log backup is not applied when the recovery model of database is "SIMPLE." In order to have the transaction log backed up, the recovery model of the database must be reset to "FULL" or "Bulk-logged."
NetWorker Module database restore process

When you start a restore, the NMSQL automatically implements the following process: 1. The NMSQL restores the most recent full backup, and then restores the most recent differential (level 1) backup (if any). If a full database backup is removed from the NetWorker server, and an incremental backup is attempted, the restore fails. The NMSQL software checks the SQL Server instance to determine if a full database backup has been performed, but does not verify that a full backup still exists on the NetWorker server. 2. The NMSQL restores all transaction log back ups that ran after the most recent differential backup (or that ran after the last full backup, if there was no differential backup). To correctly restore uncommitted transactions, the NMSQL specifies the SQL Server NORECOVERY mode for all intermediate transaction logs. The restore of the final transaction log specifies the restore mode if a mode of STANDBY or NORECOVERY was selected. The default selection is Normal.
NetWorker Module database restore process
4-7
For example, if you selected a restore mode of NORECOVERY, that specification appears in the output for a database restore as follows:
C:> nsrsqlrc -s NetWorker_server_name my_database nsrsqlrc: Restoring database my_database... nsrsqlrc: RESTORE database my_database FROM virtual_device='BSMSQL' WITH norecovery, stats nsrsqlrc: RESTORE database my_database from virtual_device='BSMSQL' WITH norecovery (differential) nsrsqlrc: RESTORE transaction my_database FROM virtual_device='BSMSQL' WITH norecovery nsrsqlrc: RESTORE transaction my_database FROM virtual_device='BSMSQL' WITH norecovery Received 1.0 MB 4 files from NSR server.
The NMSQL imposes the following restrictions on database file relocation:
Only database backups can be relocated. Individual file and filegroup backups cannot be relocated without relocating the database that contains those files. If the configuration of a database has changed since the most recent, level full database back up was created, you cannot relocate the database. Configuration changes include the deletion or addition of files, filegroups, or transaction log files. A system database might not be the destination database of a relocation. The relocation fails if the destination does not have sufficient space to create a new database.
Performing a restore with NetWorker User for SQL Server

To recover SQL Server data from a backup, perform the following tasks:

Task 1: Set up the restore on page 4-9 Task 2: Specify the browse time (optional) on page 4-33 Task 3: View the required volumes (optional) on page 4-35 Task 4: Set the restore properties (optional) on page 4-36 Task 5: Start the restore on page 4-54
4-8
Note: The SQL Server databases must be restored in the correct order.
The amount of time that it takes to restore a database depends on the following environmental variables:

The amount of data Network traffic Server load Backup volume availability Tape positioning
If the backup volume with the databases is loaded at a storage node (backup device) local to the NetWorker server, the restore proceeds. If the restore does not begin, it is possible that either the wrong volume or no volume is mounted in the backup device. When restoring an incompatible database by using the name of an existing database, or when restoring from a media failure where one or more database files were lost, the Overwrite the Existing Database attribute must be selected under the Files tab.Properties dialog box, Files tab on page 4-14 provides details. After the restore is finished, the restore completion time appears in the Restore Status window.
Task 1: Set up the restore

To set up the restore: 1. Start the NetWorker User for SQL Server program. The main window appears, as shown in Figure 2-2 on page 2-5. If multiple SQL Servers are active on the computer, the Select SQL Instance dialog box opens before the main window opens, as shown in Figure 2-1 on page 2-4. 2. Select the SQL Server host that the NMSQL should use to perform the restore operation, then click OK. The main window appears. 3. To select a NetWorker server other than the default server: a. Click the Select NetWorker Server button on the toolbar.
4-9
The Change Server dialog box appears, as shown in Figure 2-3 on page 2-6. b. Select a NetWorker server from the list, then click OK. 4. In the main window, click the Restore button on the toolbar. The Restore Operation dialog box appears.
Figure 4-1
Restore Operation dialog box Note: You can also select Restore from the Operation menu, and then select Normal, Copy, Verify Only, Partial, or Piecemeal (if running SQL Server 2005 Enterprise Edition).
Configuring a normal restore To configure a normal restore: 1. Select the Normal restore type from the Restore Operation dialog box and click Continue.
4-10
The Restore window appears listing the databases that can be restored.
Figure 4-2
Normal Restore window
2. Select one or more databases to restore. 3. Right-click each marked database and select Properties to configure additional settings. The Properties dialog box appears.
Figure 4-3
Properties dialog box, General tab
4-11
Options and actions that are available are dependant upon the version of SQL Server that is to be restored and the type of restore selected. 4. Specify the following on the General tab: Back up the active portion of the transaction log before restoring the database This option backs up the active portion of the transaction log before performing the restore. When selected, the NMSQL attempts a transaction log backup by using the NO_TRUNCATE SQL keyword for SQL Server 2000. For SQL Server 2005 databases, this option is selected by default for a Normal restore, and the NMSQL attempts the transaction log backup by using the NO_TRUNCATE and NORECOVERY SQL keywords. The restore operation proceeds regardless of whether the transaction log backup succeeds or fails.
Note: The most common reason for restoring databases is to recover from operator errors, which are recorded in the transaction log. If you recover the database without applying the transaction log you loose the information since the last backup. If you apply the entire transaction log you re-corrupt the database. With NMSQL 5.0, point-in-time recovery, data can be recovered to the time of the error minus approximately one second. This assumes that the database is functional enough to complete the final transaction log backup. If the Specify a Restore Time box under the Restore Time tab is checked, the backup proceeds, but the latest transactions captured in the active transaction log backup are not restored.
Perform the restore using this recovery mode Normal mode instructs SQL Server to leave the database in operational state after the restore. This is the default mode. No Recovery mode activates the SQL Server NORECOVERY database restore option for the last stage restored. This mode places the database in an unloadable state after the restore. However, the database can still process additional transaction log restore operations.
4-12
Standby mode specifies an undo file for SQL Server to use when rolling back the transactions. By default, this attribute displays a default filename and path:
%DriveLetter:\<default_path>\<default_dbName>undo.ldf
where: default_path is the default SQL Server backup path obtained from the SQL Server registry. default_dbName is the name of the database backup selected for the restore. To specify another name and path for the this file: a. Enter a valid name and path, or click the ellipses button. The Specify the Standby Undo File dialog box appears.
Figure 4-4
Standby Undo File dialog box
b. In the Specify the Standby Undo File dialog box, specify the following attributes: Enter a path in the File Location text box, or browse the file system tree and highlight a file. Enter the filename in the File Name text box, or browse the file system tree and highlight an existing file. Perform checksum before reading from media This option (in the Properties dialog box) performs a checksum operation before a restore to ensure that it matches the backup.
4-13
The NMSQL verifies the checksum by calculating a local result and comparing it with the stored value. If the values do not match, you can choose to continue the restore operation by selecting the Continue on checksum error option.
Note: The Checksum and Continue with error options are available starting with SQL Server 2005 instances.
5. Select the Files tab.
Figure 4-5
Properties dialog box, Files tab
Configure the following settings: Database to restore displays the database selected for the restore. This attribute cannot be modified. Name for restored database specifies the name for the restored database. For a normal restore, this text box displays the name of the database selected for backup and cannot be modified. Overwrite the existing database instructs SQL Server to create the specified database and its related files, even if another database already exists with the same name. In such a case, the existing database is deleted.
4-14
Note: This attribute includes the WITH REPLACE SQL keyword in the restore sequence. The WITH REPLACE keyword restores files over existing files of the same name and location. Microsoft SQL Server Books Online provides more information.
Mark the filegroups to restore defines the files and filegroups to restore. If performing a normal or copy restore, the filegroups of the database selected cannot be changed.
Note: The set of filegroups marked in this attribute is copied into the list of the Modify the destination for the files in attribute.
Modify the destination for the files in displays a set of views for the database files to be restored, and enables filtering of files that are visible in the File and destination table. Table 4-1 on page 4-15 identifies the supported views:
Table 4-1
Views displayed by the Modify the destination for the files in attribute This view All files All log files All data files Filegroup name Drive letter Displays All of the files for the database, including transaction log files. Only the transaction log files. Only data files. Only data files for a specific filegroup. All files located on a given drive at the time the backup occurred, even if those files have since been relocated to a different drive.
File and destination table lists the SQL Server logical filenames and locations. The files listed in this table are associated to the marked database to be restored. When performing a normal restore, this table displays the current name and destination based on the SQL Server physical filename and logical location for the restored file. To modify the destination, perform one of the following: Double-click a file in the list to display the Specify the file destination dialog box. Click a file in the list, and then click Destination to display the Specify the file destination dialog box.
4-15
Figure 4-6
Specify the File Destination dialog box
Configure the attributes listed in Table 4-2 on page 4-16.

Table 4-2
Configurable attributes This view Source file name Displays The file currently selected in the File and Destination lists. This text cannot be modified. When multiple files are selected, this text box is empty. The location and the file selected in the File and destination list. This information cannot be modified. When multiple files are selected, the location of the first selected file in the list is displayed. The file system location for the restored file. When multiple files are selected, the default SQL data path is opened, but not selected. Enter a pathname, or browse the file system tree and highlight a directory or file to change the location. The name of the file currently selected in the File and Destination table. When multiple files are selected, the attribute is empty. Enter a new name or browse the file system tree and highlight a file to change the name.
Source location
Destination location
Destination file name
6. Click OK to return to the Files tab.
4-16
7. Click the Restore Time tab to configure a restore schedule.
Figure 4-7
Properties dialog box, Restore Time tab
This tab enables you to select a backup version and modify the restore date and time. The default selection for the restore is listed in the Backup Version table. When a point-in-time restore is performed, the restore procedure reinstates only transactions from the backup version that occurred before the specified restore date and time. You can change the backup version or transaction time. The ability to restore to a named log mark is also supported for SQL Server 2000 and 2005 databases. 8. To perform a point-in-time restore, specify the following in the Restore Time tab: Select the Specify a Restore Time checkbox to schedule the restore. If the Backup the active portion of the transaction log before restoring the database checkbox on the General tab is selected, and you choose this option but do not specify the point-in-time in the log tail, the latest transactions captured in the active transaction log backup are not restored.
4-17
The latest transactions captured in the active transaction log backup will be restored to the specified point-in-time only if it was specified in the log tail. Specify a Time to Perform a Point-in-time Restore The restore time indicates what data from the marked backup version is reinstated during the restore, and when to stop restoring transactions. This text box can be modified by clicking the Point-in-Time button. Point-in-time button When the incremental backup, or the latest backup (regardless of whether it is full, level incremental or level differential) is marked in the Backup Versions table, the Point-in-time button is enabled for setting the time for a point-in-time backup. Set point-in-time restore properties on page 4-52 provides more details. If a full or differential backup is marked but is not the latest backup, this button displays an error message. Backup Versions table Select and mark the backup version to use for the restore. The Specify a Time text box displays the date and time for the backup that is currently marked. Only one backup version can be selected and marked. To select a backup version, double-click a backup or select a backup and click the Mark button. Using a Named Log Mark Perform the restore by using a named log mark. This attribute: Is only available for SQL Server 2000 and 2005. Is only enabled when log marks exist for the selected database backup. When this attribute is selected, the Restore to the End of the Log Mark and the Restore to the Beginning of the Log Mark buttons are enabled.
4-18
Specify which type of named log mark restore to perform by selecting one of the following: To restore the backup and stop immediately after the named log mark, select Restore to the End of the Log Mark. This type of restore includes the named transaction in the restore. To restore the backup and stop immediately before the named log mark, select Restore to the Beginning of the Log Mark. This type of restore excludes the named transaction. Log Mark table Use this attribute to specify a log mark to use for the restore. Double-click a log mark or select a log mark in the list and click the Mark button. 9. When finished, click OK. There are additional restore options that you can define. 10. Select Restore Options from the Options menu. The Restore Options dialog box appears.
Figure 4-8
Restore Options dialog box, General tab
4-19
11. From the General tab, configure the following attributes: Automatically overwrite any database marked for restore Select this attribute to overwrite the marked databases. The NMSQL restores the database by using the WITH_OVERWRITE SQL keyword. The following message dialog box prompts you to confirm the action.
Detect available tape sessions prior to restoring a striped backup Select this attribute to restore SQL data that was originally backed up as a striped backup. If this attribute is selected, the NMSQL determines the number of tape sessions needed to restore a striped backup. The following message dialog box prompts you to confirm the action.
Note: This attribute is selected by default. It is controlled by the Windows registry entry NSR_DETECT_TAPES, which can be modified. Regardless of whether this attribute is selected, it retains its setting from one session to the next. Appendix C, Striped Backup and Recovery, provides more information on striped backups and this registry keyword.
Note: If you do not have adequate permissions, this checkbox is disabled. Security requirements on page 1-2 provides information about how to determine the permissions necessary for Windows 2000 servers. Appendix B, NetWorker Module Permissions Wizard, provides information about how to determine the permissions necessary for Windows 2003 servers.
4-20
Select a Debug Level Selecting a debug level issues the command line D n option, where n is an integer value between 1 and 9. When the debug level is selected, the debug information appears in the Backup or Restore status window, and also in the log files. This information can be useful to assist Tech Support in diagnosing problems. If you specify debug level logging, watch the size of the module log files. The files are located at: installpath\nsr\applogs Where: installpath is typically C:\Program Files\Legato and can grow very large. Log files with debug output can be deleted (or archived) once EMC Tech Support has a copy. Disable Log Mark Display This attribute disables the potentially time-consuming retrieval of log mark information. Select in situations where log marks will not be used for most restore operations, particularly if the client file index is large. This setting is remembered across invocations of the user interface. When log mark display is disabled, the log mark in the backup will not be displayed in the Restore Time tab of the Restore Properties dialog box. When this option is selected, the log mark information is not displayed in the Restore Time tab of the Restore Properties dialog box. When it is not selected, the information appears in the Restore Time tab of the Restore Properties dialog box. Log Tail backup before restoring Microsoft SQL Server 2005 databases (SQL Server 2005 only) With SQL Server 2005, transaction log backups are required prior to restoring databases. This ensures that the database can be restored to the most recent point-in-time possible. NMSQL 5.0 selects this option by default for SQL Server 2005. When this option is not selected, the Back Up the Active Portion of the Transaction Log Before Restoring the Database option (selected by default for SQL Server 2005 databases) is enabled. By deselecting the Back Up the Active Portion of the Transaction Log Before Restoring the Database option, the following shown in Figure 4-9 on page 4-22 appears.
4-21
Figure 4-9
Review Options message
You can choose not to back up the transaction log by selecting No for Restore with overwrite in the precautionary message box. This will replace the existing database by using the T-SQL option WITHREPLACE. This will improve performance in cases where the database is too damaged to perform the transaction log backup or it is known that the transaction log is not of interest. 12. (Optional) If the PowerSnap Module is installed, select the Snapshot tab to configure snapshot restore options. Choose from either File Level or Volume Level rollback operations. Other file system objects on the affected volumes will be reverted to their state at the time of the original snapshot (out of date or nonexistent). Other databases that reside partially, or in total, on the affected volumes are not synchronized at the time of the original snapshot and will be out of date and possibly corrupt after a rollback. If a rollback fails, the database and snapshot backup are lost. The PowerSnap documentation provides information on how to specify file system objects that should be ignored during volume rollback safety checks. These safety checks can be overridden by selecting the Force rollback checkbox.
4-22
An additional dialog box prompts users to continue or cancel if other database files are detected on the volumes.
Figure 4-10
Restore Options dialog box, Restore tab
File Level Restore Type If you select File Level restore, then a file level restore procedure runs from the source selected in the Restore Type windows. Volume Level Rollback If you select the Volume Level Rollback button, then file-level restore is not performed. Restore Type and Restore Type Order Use the right and left arrows to select one or both of the snapshot restore methods, and use the up and down arrows to establish the order of procedure: Persistent Snapshot If only the persistent snapshot option is selected, a snapshot is restored from the SQL Server storage subsystem. If a recoverable snapshot is not found on the SQL Server storage subsystem, no recovery is performed.
4-23
NetWorker Server If only the NetWorker Server option is selected, a snapshot is restored from secondary storage on the NetWorker server or storage node. If a recoverable snapshot is not found on secondary storage, no recovery is performed. Persistent Snapshot, NetWorker Server If both options are selected, with Persistent Snapshot first, a snapshot is restored from the SQL Server storage subsystem. If a recoverable snapshot does not exist on the SQL Server storage subsystem, a secondary storage search is performed on the NetWorker server or storage node. NetWorker Server, Persistent Snapshot If both options are selected, with NetWorker Server first, a snapshot is restored from secondary storage on the NetWorker server or storage node. If a recoverable snapshot is not found on the NetWorker Server or storage node, the SQL Server storage subsystem is searched for a recoverable snapshot. Specify a Data Mover This attribute identifies the hostname of the computer that restores the snapshot. By default, this value is set to the SQL Server host. The data mover information is ignored for volume-level rollback. For file level restore, enter the data mover in the Specify a Data Mover text box.
Note: If a data mover other than the target SQL Server was specified for the snapshot backup, the same data mover must be specified for snapshot recovery. The NetWorker PowerSnap Module Installation and Administrators Guide provides more information about configuring a data mover.
13. Click OK. 14. Click the play button in the toolbar of the restore dialog box to start the restore. You can monitor the status of the restore in the Restore Status window.
4-24
Configuring a copy restore To configure a copy restore: 1. Select the Copy restore type from the Restore Operation dialog box and click Continue. The Select the SQL Server dialog box appears.
Figure 4-11
Select the SQL Server dialog box
2. Specify the NetWorker client from which SQL Server backups are imported for restore to the local SQL Server by selecting one of the following attributes: The current SQL Server Instance Displays the name of the current NetWorker client host. Select this attribute to use a database backup from the SQL Server currently connected to the NMSQL. A [different] NetWorker SQL Server client By default, this attribute is disabled and the text box displays the name of a NetWorker client host. Select this attribute to select a different SQL Server host from which to obtain the existing database backup.
Note: In order to select other client machines as the source of a copy restore, the destination machine and an appropriate user account or group must be specified in the list of NetWorker administrators on the NetWorker server. The release notes provide information about setting up a mid-level user account that does not grant full administrator privileges.
4-25
A new dialog box appears if the following applies to the selected NetWorker client: A backup exists. A named instance directory in the index. This dialog box enables the selection of the source from default and named instances. 3. Select either the SQL Server default instance or an SQL Server named instance, and click Continue. After choosing an SQL instance, all database backups are listed.
Note: CopyOf is prepended to database (and database file) names only if a database by the same name already exists on the target system. This may not be the case for copy restore between systems or between instances of SQL Server (or if the original database has been removed). In these cases the default database and filenames will be the same as the original names.
If backups for the SQL default or other named instance were not performed, or are no longer available, the message shown in Figure 4-12 appears.
Figure 4-12
NetWorker User for SQL Server message
4. After you select a SQL instance, the Restore window appears listing the databases that can be restored.
4-26
Figure 4-13
Restore (Copy) window
5. Select one database to restore.

Note: If the partial, piecemeal, or copy restore type is selected, only one database object can be marked. Files and filegroups of the selected database are not visible in the browse window since they require a database backup. The files and filegroups can be viewed and configured in the Files tab of the Properties dialog box.
6. Right-click the marked database and select Properties.

Note: If you select partial, piecemeal, or copy restore type, you must specify the restore properties before starting the restore. Otherwise, the restore fails to start.
The Properties dialog box appears and the General tab is displayed.
Figure 4-14
Sample Properties dialog box
4-27
7. In the Properties dialog box, click the Files tab. When restoring a copy of a database, you can overwrite an existing database or create a new one. To create a new database, enter a name in the Name for restored database text box. The default is CopyOfSelectedDB. If you want to overwrite an existing database, select a name from the Name for restored database text box and select Overwrite the existing database.
Note: Check the generated filenames. If you change the name of the database from CopyOfAcme Sales back to Acme Sales, by using overwrite, then you may remove the CopyOf from the associated filenames as well. Default names are generated when the dialog box is first displayed. Any database name edits are not propagated to the database filenames. These changes must be made manually.
8. In the Properties dialog box, click the Restore Time tab. 9. When properties are configured, click OK. These settings are maintained until the restore is started. 10. Click the play button in the toolbar of the Restore dialog box to start the restore.
Note: You can monitor the status of the restore in the Restore Status window.
To perform a copy restore of a snapshot backup, these requirements must be met:
The NetWorker servers file on the data mover node (proxy client) must contain an entry for the destination client. The NetWorker Administrators Users group must include system@destination_client_hostname (using a fully qualified domain name) in the Users attribute. The snapshot backup must have been rolled over to tape (or other file device). Copy restore does not work with volume rollback.
Named instance recovery For a recovery, a default instance and a named instance A and B are available on a client computer. Once a backup is performed by using NetWorker User for SQL Server, a copy restore is available for all three instances.
4-28
If instance A is uninstalled for some reason, a copy restore from instance A is still available, because the instance is still listed in the client file index for the most recently named instance directory. Should a more recent backup be performed for any of the instances, after instance A is uninstalled, a recovery will not be possible. A new named instance directory is created with instance A excluded. The remedy for this is to reinstall the removed named instance and run another backup. This restores the directory information for instance A and complete functionality.
Figure 4-15
Restore window
1. To view a list of data items available for restoring, expand any item in the left pane of the Restore window. The descendants of the item are displayed in the right pane. 2. In the Restore window, select the item to restore by clicking the checkbox.
Note: If the partial, piecemeal, or copy restore type is selected, only one database object can be marked. Files and filegroups of the selected database are not visible in the browse window, and require a database backup. The files and filegroups are available in the Properties dialog box.
4-29
3. Once the restore is set up, start the operation. Task 5: Start the restore on page 4-54 provides more information. Configuring a verify-only restore The verify-only option enables you to verify that data for a database can be restored.
Note: Verify-only restore will not work with PowerSnap PIT backups. For verify-only restore to work, the backup must have been rolled over to tape or some other device.
To configure a verify only restore: 1. Select the Verify Only restore type from the Restore Operation dialog box and click Continue. The Restore window appears listing the databases that can be restored.
Figure 4-16
Verify Only Restore window
2. Select one or more databases from the list. 3. (Optional) Right-click the parent server to open the Restore Options dialog box. 4. Click OK. 5. Click the play button in the toolbar of the Restore dialog box to start the restore. You can monitor the status of the restore in the Restore Status window.
4-30
Configuring a partial or piecemeal restore Partial restore is enabled for SQL Server 2000 releases. If you are restoring SQL Server 2005 databases (Enterprise Edition only), the option becomes piecemeal restore. Restore types on page 1-15 provides detailed descriptions. To configure a partial or piecemeal restore: 1. Select the Partial (or Piecemeal) restore type from the Restore Operation dialog box and click Continue. The Restore window appears.
Figure 4-17
Partial Restore (or Piecemeal) window
2. Select a database from the list.

Note: In partial and piecemeal restore, only one database object can be marked. Files and filegroups of the selected database are not visible in the browse window. These operations require a database backup. The files and filegroups can be viewed and configured in the Files tab of the Properties dialog box.
3. (Optional) Right-click the parent server to open the Restore Options dialog box. 4. Click OK. 5. Right-click the database and select Properties.
Note: In partial and piecemeal restore operations, you must specify the restore properties before starting the restore. Otherwise, the restore fails to start.
4-31
The Properties dialog box, similar to that shown in Figure 4-14 on page 4-27, appears. In the General tab, partial and piecemeal restores overwrite a database when restoring to the same location. To overwrite an existing database, select a name from the Name for restored database text box and select Overwrite the existing database. 6. In the Properties dialog box, click the Restore Time tab. 7. When finished configuring properties, click OK. Settings in this dialog box are maintained until the restore is started. For piecemeal restore, certain settings in the Properties dialog box are reset if you redisplay the Properties dialog box again before starting the restore process. The selections that will be reset are: Marked files and filegroups of the selected database, Name for restored database option (on the Files tab), Backup the active portion of the transaction log before restoring the database checkbox (on the General tab). Redisplaying the Properties dialog box in piecemeal restore causes the previous selections of these options to be removed. If you do return to the Properties dialog box, the message shown in Figure 4-18, appears:
Figure 4-18
Restore Options message
8. Click the Play button in the toolbar of the Restore dialog box to start the restore.
Note: You can monitor the status of the restore in the Restore Status window.
4-32
Note: Piecemeal restore is iterative. You can continue to restore additional filegroups in subsequent operations. Previously restored filegroups will not be available for selection unless you specify New Piecemeal.
Task 2: Specify the browse time (optional)

In the NetWorker User for SQL Server Restore window, you can browse the online index and mark a database backup version to restore. The browse time controls which backup data is viewable in the Restore window. You can modify the browse time to display backup versions for a different restore time by selecting Change Browse Time from the View menu. However, the browse time cannot change to a point:

Earlier than the first SQL Server module backup. Later than the most recent backup.
An invalid time entry results in the following error message:
To specify a browse time: 1. From the View menu, select Change Browse Time.
4-33
The Change Browse Time dialog box appears.
Figure 4-19
Change Browse Time dialog box
2. Select one of the following attributes to change the browse time: Specify a browse time By default, this attribute displays the current browse time. Select this attribute to enable the Browse Time button. a. Click the Browse Time button to open the Specify Browse Time dialog box.
Figure 4-20
Specify Browse Time dialog box
4-34
b. Select a date and time from the Date and Time lists. c. Click OK. The dialog box closes and the Specify a Browse Time attribute in the Change Browse Time dialog box appears with the new date and time. The backup version list is also updated. Select a Backup Version to Change Browse Time By default, this attribute marks the current browse time in the backup version list. Select this attribute to enable the backup version list and Mark button. The backup version list displays the known backup version times, which correspond to save times of the root browse directory. The root browse directory is created anytime a SQL Server backup is performed. To select a new browse time, do one of the following: Double-click a backup time in the list, and then click OK. Select a backup time in the list, click Mark, and then click OK. The Specify a Browse Time attribute is updated to reflect the new setting.
Note: When you change the browse time, previous selections in the Restore window are discarded. Any file, filegroup, or database property settings are lost.
Task 3: View the required volumes (optional)

After you select the databases to restore, you can check which backup volumes contain the databases you need by doing one of the following:

Selecting Required Volumes from the View menu. Selecting the Required Volumes button from the toolbar.
Checking the required volumes helps to ensure that the necessary backup volumes are mounted in the NetWorker server's storage device.
Note: Required volumes information is not available for snapshots.
4-35
If a required backup volume is not mounted, do one of the following:
Contact the NetWorker administrator and request that the volume be mounted. Wait for the NetWorker software to prompt the administrator to mount the volume.
Task 4: Set the restore properties (optional)

To set the restore properties for a select file, filegroup, or database: 1. Right-click the database item, and select Properties. 2. Complete the Properties dialog box. 3. Click OK.
Note: Restore properties are not available for the verify-only restore type.
See these sections for additional information:
Set database restore properties on page 4-37 provides information about setting restore properties for a database. Select filegroups to restore on page 4-42 provides information about setting restore properties for a file or filegroup.
All of the properties are context-sensitive, based on the version of SQL Server that is running (7.0, 2000, or 2005) and the restore type that was chosen. Preferences specified for these properties are used for the current restore operation. When a preference is not set, the NMSQL uses the default. After the restore is complete, and the Restore window is closed, the property settings revert back to the NetWorker User for SQL Server defaults.
Note: If you select partial, piecemeal, or copy restore type, you must specify the restore properties before starting the restore. Otherwise, the restore fails to start.
4-36
Set database restore properties The information in this section applies to normal, copy, and piecemeal restore types. To set database restore properties: 1. In the Restore window, select a database 2. Right-click the database and select Properties. The Properties dialog box appears.
Figure 4-21
Properties dialog box
3. On the General tab, indicate whether to back up the active portion of the transaction. When selected, the NMSQL attempts a transaction log backup by using the NO_TRUNCATE SQL keyword (for SQL Server 2000) and also the NORECOVERY keyword (for SQL Server 2005). The restore operation proceeds regardless of whether the transaction log backup succeeds or fails. You can specify a point-in-time restore to any time within the log tail backup. Option applies to both SQL Server 2000 and SQL Server 5000 restores, and addresses database failure if that occurs after the last scheduled incremental backup.
4-37
By backing up the active portion the active portion of the log prior to the restore, users can specify a restore for time of failure. 4. To specify a point-in-time restore, select the Restore Time tab, and make entries as appropriate. Point-in-time restore can use either a full, level incremental, or level deferential backup, as long as the selected backup is an incremental and/or the latest backup. 5. Specify a recovery mode to use for the restore: Normal mode instructs SQL Server to leave the database in operational state after the restore. This is the default mode. No Recovery mode activates the SQL Server NORECOVERY database restore option for the last stage restored. This mode places the database in an unloadable state after the restore. However, the database can still process additional transaction log restore operations. Standby mode enables the Standby Undo File Name which specifies an undo file for SQL Server to use when rolling back the transactions. (See Specifying an undo filename on page 4-38 for instructions.) By default, this attribute displays a default filename and path as follows:
%DriveLetter:\<default_path>\<default_dbName>undo.ldf
where: default_path is the default SQL Server backup path obtained from the SQL Server registry. default_dbName is the name of the database backup selected for the restore. 6. Select Checksum and Continue with error. If checksums are not calculated for the backup specifying them for the restore causes the restore operation to fail. Specifying an undo filename To specify the undo filename, do one of the following: 1. Enter a valid name and path in the text box. Click the ellipses button to display a dialog box similar to that shown in Figure 4-22 on page 4-39. 2. Select a file from the file system tree. The file location and name appear in the text boxes.
4-38
Figure 4-22
Specify the Standby Undo File dialog box
3. Select a file from the file system tree. The file location and name appear in the text boxes. 4. Specify these attributes: File Location Displays the path for the standby undo file. Enter a path in the File Location text box, or browse the file system tree and highlight a file. When a file in the browse tree is highlighted, the files path appears in the File Location text box. File Name Specifies the standby undo filename. Enter the filename in the File Name text box, or browse the file system tree and highlight an existing file. When a file is highlighted, the filename appears in the File Name text box. d. Click OK. View file or filegroup restore properties The file and filegroup restore properties in the Properties dialog box are provided for informational purposes only. SQL Server dictates the settings for these properties. They cannot be configured. The properties in these dialog boxes apply to one selected or a filegroup for a normal restore type operation.
4-39
Note: These properties are not available for partial, copy, or verify-only restore type operations.
To view the restore properties for a file or filegroup: 1. From either pane of the Restore window, right-click a selected file or filegroup. 2. Select Properties. The Properties dialog box appears. Properties differ depending on the version of SQL Server that is run. Figure 4-23 on page 4-40displays the properties for a filegroup.
Figure 4-23
Filegroup Restore Properties dialog box
4-40
Figure 4-24 on page 4-41displays the properties for a file.
Figure 4-24
File Restore Properties dialog box
The following attributes appear in the Properties dialog box: Backup the active portion of the transaction log file When selected the active portion of the transaction log is backed up before performing a restore. That way, the log can be applied to the filegroup or file to make it consistent with the rest of the database. The SQL Server requires the transaction log when restoring damaged or lost data files. NetWorker User for SQL Server attempts a transaction log backup as follows: For versions prior to SQL Server 2005, the backup uses the NO_TRUNCATE SQL keyword. The restore proceeds regardless of whether the backup was successful. For SQL Server 2005 non-Enterprise Editions or 2005 Primary filegroup, the backup uses the NO_TRUNCATE and NO_RECOVERY SQL keywords. For files belonging to secondary filegroup and secondary filegroups restore for SQL Server 2005 Enterprise Edition, the restore workflow requires you to first restore the filegroup and
4-41
then take a backup of the active portion of the transaction log. The transaction log backup must be applied to the file or filegroup restore to ensure that the file or filegroup is consistent with the rest of the database. If a file or filegroup is restored with the NetWorker User for SQL Server program, these transaction log backups occur automatically. It is recommended that you use the NetWorker User for SQL Server for this type of restore. Overwrite the existing filegroup/file with the restored file Forces SQL Server to ignore errors due to nonexistent files which result from media failure. If there is a media failure, then the files cease to exist. The NetWorker User for SQL Server specifies the WITH REPLACE SQL keyword in the restore sequence. The file or filegroup is restored to the exact location (drive and pathname) as the location on the source host from which the data was backed up. Backup versions table Lists the date and time of all the backups available for the restore operation. Select filegroups to restore Use the Properties dialog box to select a filegroup to restore. Tabs appear differently depending on the type of restore:
For normal and copy restore, the tab is labeled Files and is supported for SQL Server versions 7.0, 2000, and 5000. For a partial restore, the tab is labeled General and is available only for SQL Server 2000. For a piecemeal restore, the tab is labeled Files and is supported only for SQL Server 2005.
Note: If the marked database item selected was created by a release of the NMSQL earlier than 3.0, or the most recent backup is a transaction log backup for a database that was corrupt, a Files tab selection may first open the Read File Configuration dialog box. Specify Read File Configuration properties on page 4-47 provides further details.
4-42
To select filegroups to restore: 1. Select the Files tab.
Figure 4-25
The Files tab of the Properties dialog box
2. Specify attributes as follows:

Note: If the text boxes in this dialog box are empty, review the file configuration information. For further details, see Specify Read File Configuration properties on page 4-47.
Database to restore Displays the name of the database (on secondary storage) selected for the restore. This attribute is informational only and cannot be modified. Name for restored database Specifies the name for the restored database: If performing a normal restore, this text box displays the name of the selected database is disabled.
4-43
If performing a partial or copy restore, the NMSQL displays the default name by appending CopyOf or PartOf to the source database name, and to all associated data files and log files. To specify a different name, enter a new name in the text box or select a name from the list. The name must comply with SQL Server naming conventions.
Note: If you specify a different name, the data and log files retain the default name, as shown in the File and Destination table. For example, if copy restore is selected when restoring a database named Project to a database named Test, and the data and log filenames retain the values of CopyOfProject_Data.MDF or CopyOfProject_Log.LDF. The data and log filenames must be changed. Specify the restored files destination and filename on page 4-46 provides information to change data and log filenames.
When the Name for restored database attribute is set to the name of an existing database, the Overwrite the existing database attribute is enabled when you click Apply or OK. These two attributes can then be used together. The name of the existing database is then used for the restored database when the two databases are incompatible. Overwrite the existing database Instructs the SQL Server to create the specified database and its related files, even if another database already exists with the same name. In such a case, the existing database is deleted.
Note: This attribute causes the WITH REPLACE SQL keyword to be included in the restore sequence. The WITH REPLACE keyword restores files over existing files of the same name and location. The Microsoft SQL Server Books Online provides more information on the WITH REPLACE SQL keyword.
Mark the filegroups to restore Select or clear the filegroups to restore when the following applies: If performing a normal or copy restore this attribute displays the filegroups of the database selected. If performing a partial or piecemeal restore, by default, this attribute displays the filegroups of the database marked for the restore.
4-44
To select or deselect a filegroup: a. Highlight the filegroup in the list. b. Click the Mark/Unmark button. You can select multiple filegroups. In SQL Server 2000, the primary filegroup is always marked and cannot be unmarked. SQL Server requires that the primary filegroup be included in a partial restore. In SQL Server 2005, the primary filegroup is always marked in the initial stage of a piecemeal restore, and cannot be unmarked. Note that the piecemeal restore is iterative. You can continue to restore additional filegroups in subsequent operations. Previously restored filegroups will not be available for selection unless you specify New Piecemeal.
Note: The set of filegroups marked in this attribute is copied into the Modify the Destination for the files in attribute list.
Modify the destination for the files in This list contains a set of different views for the database files to be restored, and enables filtering of files that are visible in the File and Destination table. The views listed in Table 4-1 on page 4-15 are supported. File and Destination table The tabless File column lists SQL Server logical filenames. The Destination column lists physical filename and locations. The files listed in this table are associated to the marked database to be restored. If performing a normal restore, this table displays the current name and destination based on the SQL Server physical filename and logical location for the restored file. If performing a partial or copy restore, this table displays a default name and destination based on the SQL Server physical filename and logical location for the restored file.
Note: The default location for the data files and log files is in the data path of the default SQL Server installation directory. If this directory is on the system drive, provide enough disk space for the database files, or specify another location that does.
4-45
You cannot edit the File and Destination table. You can, however, modify the destination location. To modify the destination, do one of the following: Double-click a file to display the Specify the file destination dialog box, as shown in Figure 4-26 on page 4-46. Then follow the instructions in the next section. Click a file, and then click the Destination button to display the Specify the file destination dialog box. Then follow the instructions in the next section. Specify the restored files destination and filename Specify the destination locations for the restored files in the Specify the File Destination dialog box.
Figure 4-26
Specify the File Destination dialog box
Specify attributes as follows:
Source file name displays the file currently selected in the File and Destination lists. The Source File Name text box is informational only and cannot be modified.When multiple files are selected, this text box is empty. Source location displays the file system location and the file currently selected in the File and Destination lists. The Source Location text box is informational only and cannot be modified.
4-46
When multiple files are selected, this text box contains the file system location of the first selected file in the File and Destination lists.
Destination location displays the file system location for the restored file. When multiple files are selected, the default SQL data path is opened, but not selected. To modify this attribute enter a pathname, or browse the file system tree and highlight a directory or file. When a directory is highlighted, that path appears in the Destination Location text box. If a file is highlighted, the directory for the highlighted file is displayed.
Destination file name, by default, lists the name of the file currently selected in the File and Destination table. When multiple files are selected, the attribute is empty. To modify this attribute, enter a new name in the Destination File Name text box or browse the file system tree and highlight a file. When a file is highlighted, the filename is displayed in the Destination File Name text box.
Note: Default filenames are generated when the dialog box is first displayed. Verify that the filenames are correct. This is particularly important after changes to the database name.
Specify Read File Configuration properties Some of the data used to populate the attributes on the Files tab of the Properties dialog box is obtained from new file-configuration metadata objects created in the client file index. For backups created with a release earlier than 3.1, the file-configuration metadata is not present in the client file index, but is available in the save set media.
Note: NMSQL release 3.0 transaction log backups may not create metadata.
To specify Read File Configuration properties: 1. Open the Properties dialog box for a marked database item that has no file-configuration metadata in the client file index. You may load this information from the save set media.
4-47
The Read File Configuration dialog box appears.
Figure 4-27
Read File Configuration dialog box
2. Specify attributes as follows: Read the file configuration from save set media If you read from tape, a message appears as shown in Figure 4-28. The save set media is read in the background. When this process finishes, the Properties dialog box appears, and the Filegroup and Destination table display valid data.
Figure 4-28
Reading dialog box
To cancel the reading process, select the Cancel button. If you cancel from the Reading dialog box, the Properties dialog box appears, but the File and Destination table is empty. Skip reading the file configuration If you select this attribute, the Properties dialog box appears, but the File and Destination table is empty. Set Restore Time properties The Restore Time properties support selection of a backup version and modification of the restore date and time. When a point-in-time restore is performed, the restore procedure reinstates transactions only from the backup version that occurred before the specified restore date and time.
4-48
The ability to restore to a named log mark is also supported for SQL Server 2000 and SQL Server 2005. To set restore time properties: 1. In the Properties dialog box, select the Restore Time tab.
Figure 4-29
The Restore Time tab in the Properties dialog box Note: If the back up the active portion of the transaction log checkbox on the General tab is selected, and you select the Specify a Restore Time attribute, the transactions in the log tail backup will not all be present in the restored database. Only up to the point-in-time restored appear. You can specify a point-in-time restore between the last scheduled full, incremental, or differential backup and the current time. The transaction log will be backed up as the initial portion of the restore operation and, if the transaction log backup is successful, the log is also restored.
2. Select Specify a Restore Time to enable these restore time controls: Using a specific backup version When this attribute is selected, the following are enabled.
4-49
Specify a time to perform a point-in-time restore Indicates what data from the marked backup version is reinstated during the restore, and the time to stop restoring transactions. This text box can be modified for a backups selected in the Backup Versions table if a backup is the latest or an incremental backup. Point-in-time button If an incremental backup is selected in the Backup Versions table, that text box (Figure 4-29 on page 4-49) can be modified by clicking the Point-in-Time button. The Point-in-Time button is also enabled for any latest backup (full, incremental, or differential) for which the Backup the active portion of the transaction log before restoring the database checkbox is checked on the General tab. For details, see Set point-in-time restore properties on page 4-52. If the backup that is marked is not the latest backup, this button displays an error message. Backup versions table Select and mark the backup version to use for the restore. The Specify a Time text box displays the date and time for the backup that is currently marked. Only one backup version can be selected and marked. The Backup Versions table contains the four columns identified in Table 4-3 on page 4-50.
Table 4-3
Backup Versions table columns This column Size Type Method Displays The size of the backup. The backup type, including full for level full backups, incr for level incremental backups, and 1 for level differential backups The method, either Traditional Recover or Snapshot Recover, used to originally create the backup that is to be restored. The Snapshot Recover method is only available if you select the normal restore type. The date and time, in seconds, when the backup was created
Backup Time
4-50
To select a backup version, do one of the following: Double-click any backup in the table. Click any backup in the table, then click the Mark/Unmark button. When a backup version is marked, a check mark appears next to the backup size. If another backup version is already marked, the newly selected backup version is marked and the other version becomes unmarked. Using a named log mark Perform the restore by using a named log mark. If log marks are not used, then display of this information can be disabled. This attribute has these characteristics: Can use a maximum of 1024 log marks. Is visible only in SQL Server 2000 and 2005. Is enabled only when log marks exist for the selected database backup. When this attribute is selected, the Restore to the End of the Log Mark and the Restore to the Beginning of the Log Mark buttons are enabled. Specify which type of named log mark restore to perform by selecting one of the following: To restore the backup and stop it immediately after the named log mark, select Restore to the End of the Log Mark. This type of restore includes the named transaction in the restore. To restore the backup and stop it immediately before the named log mark, select Restore to the Beginning of the Log Mark. This type of restore excludes the named transaction. Log Mark table Use this attribute to specify the log mark to use for the restore. Only one log mark may be selected.Table 4-4 on page 4-52 shows the columns in the Log Mark table.
4-51
Table 4-4
Log Mark table columns This column Log mark Date Description Displays The name of the transaction log mark. The date and time, in granularity of milliseconds, on which the named transaction log mark was created. The any information about the log mark that was entered when the transaction was performed.
To select a log mark do one of the following: Double-click any log mark in the table. Click any log mark in the list, then click the Mark button. A check mark appears next to the log mark name. If another log mark is already selected, the newly selected log mark is marked and the other log mark becomes unmarked. Set point-in-time restore properties Specify a date and time for the SQL point-in-time restore in the Point-in-Time Restore dialog box. Be sure that the date and time are within a timeframe spanned by the transaction log backup that was marked in the Backup Versions table. The create time of the transaction log backup defines the upper limit of the time frame. No date and time greater than this upper limit can be specified. If a transaction log backup has been specified as part of the restore, , then the upper limit is the current time. The lower limit of the time frame is the create time of the most recent level incremental or level full backup created prior to the selected backup. No date and time less than or equal to this lower limit can be specified. To set point-in-time restore properties: 1. On the Restore Time tab: a. Select the Specify a restore time checkbox. b. Click Using a specific backup version button. c. Select the Point-in-time button.
4-52
The Point-in-time restore dialog box appears.
Figure 4-30
Point-in-Time Restore dialog box
2. Specify the following attributes as needed: For Date, specify the restore time date for the marked backup version. The syntax for this text box is mm/dd/yyyy. To modify the date, enter a date by using the appropriate syntax or click the arrow to display the Calendar. In the Calendar, click a date. Use the Previous Month and Next Month buttons to change from the current month. The restore time date must be within the range of transaction dates included in the selected backup version. If a transaction log backup has been specified as part of the restore operation then the upper limit is today's date. For Time, specify the restore time for the marked backup version. The syntax for this text box is hh:mm:ss. To modify the time, enter a time by using the appropriate syntax or use the scroll arrow buttons to change the time. The restore time must be within the range of transaction times included in the marked backup version. If a transaction log backup has been specified as part of the restore operation then the upper limit is the current time.
4-53
Task 5: Start the restore

To start a restore: 1. Click Start to begin the restore. 2. To cancel the restore, select End Restore from the File menu. The amount of time to restore databases depends on:

The amount of data Network traffic Server load Backup volume availability Tape positioning
While the restore is in progress, the Restore Status window displays messages for each restored database to monitor the progress of the restore. If the backup volume containing the databases is loaded at a storage node (backup device) local to the NetWorker server, the restore proceeds. If the restore does not begin, it is possible that either the wrong volume or no volume is mounted in the backup device. When restoring an incompatible database by using the name of an existing database, or when restoring from a media failure where one or more database files were lost, the Overwrite the Existing Database attribute must be selected under the Files tab. To display the Files tab, right-click the database item and select Properties. After the restore is finished, the restore completion time appears in the Restore Status window.
4-54
5
Invisible Body Tag
This chapter explains the NetWorker Module for Microsoft SQL Server (NMSQL) backup and restore procedure for a Microsoft Cluster Server (MSCS) environment, and how this procedure relates to MSCS cluster support provided for NetWorker file system clients. This chapter includes the following sections:

How the NetWorker client works in a Microsoft Cluster Server 5-2 How the NMSQL works in a Microsoft Cluster Server................5-3 How to run a scheduled backup......................................................5-5 How to run manual backups and restores for a clustered SQL Server .................................................................................................5-15
5-1
How the NetWorker client works in a Microsoft Cluster Server

An MSCS cluster is a set of nodes and virtual servers.
On Windows 2000 Advanced Server, MSCS supports two-node clusters. On Windows 2000 Data Center, MSCS supports up to four-node clusters. On Windows Server 2003 (Enterprise or Datacenter Edition), MSCS supports up to eight-node clusters.
Microsoft Knowledge Base article: 288778, Maximum Number of Supported Nodes in a Cluster provides more information on the number of nodes supported by MSCS. Each node is a physical computer with its own IP address and network name. A cluster may be configured to contain any number of virtual servers. Each virtual server :

Has its own IP address and network name. Owns a subset of shared cluster disks Is responsible for starting cluster applications that can fail over from one cluster node to another.
The NetWorker client software regards each physical node and each virtual server in a cluster as a separate NetWorker client. Each NetWorker client has its own client file index on a NetWorker server, and each client is responsible for backing up its own files.
A client associated with a physical node in the cluster backs up files on private disks attached to that node. A client associated with a virtual server backs up only files on disks in the cluster resource group belonging to that virtual server.
Note: To back up all of the cluster's data, both the NMSQL software and the NetWorker client software must be installed on a private disk on each physical node in the cluster. The NetWorker Module for Microsoft SQL Server Installation Guide provides information about installing the NMSQL to run in a cluster environment.
5-2
How the NMSQL works in a Microsoft Cluster Server

The NMSQL can back up or restore data exported by SQL Server that is running as a virtual server in an MSCS cluster. The NMSQL uses the virtual server name to do the following:

Connect to the appropriate SQL Server. Initialize the SQL Server VDI to accept data from, or deliver data to, the appropriate SQL Server in the cluster. Create entries in the NetWorker client file index.
The NetWorker media database or client file index is indexed according to the client that performs a particular backup. The NMSQL creates entries under the virtual server name in the NetWorker client file index. Information about the SQL Server save sets of a virtual server is stored in the NetWorker client file index associated to that virtual server name.
Note: To use the NMSQL to back up and restore SQL Server data on a node in a cluster, you must have Cluster Client Connection licenses on the NetWorker server host (one for each node in the cluster). NetWorker Power Edition includes two cluster client licenses of the same platform type. For more than two cluster nodes, additional Cluster Client Connection licenses are required. NetWorker Workgroup Edition and NetWorker NetWork Edition can back up cluster nodes only if Cluster Client Connection licenses have been added to the NetWorker server.
How the NMSQL detects SQL Server instances

When running in an MSCS cluster, the NMSQL automatically detects all active SQL Servers in the cluster, including virtual servers. This automatic detection occurs whenever:

The NetWorker User for SQL Server program is started. The Select SQL Instance menu item or button is selected. A backup or restore is started.
How the NMSQL works in a Microsoft Cluster Server
5-3
Named instances in failover cluster configurations

The NMSQL provides failover cluster support by using the multiple instance features provided in SQL Server 2000 and 2005. In a failover configuration, virtual servers run as either the default instance or as named instances. One default instance of a virtual server may be installed. Additional virtual servers may be installed as named instances, where each instance name must be unique within the cluster. The NMSQL supports multiple named instances as virtual servers in a cluster configuration. The number of instances supported depends on the version of SQL Server being used. The NetWorker Module for SQL Release 5.0 Installation Guide and Microsoft SQL Server documentation provides more information. Each named instance virtual server has the following qualities:

A unique IP address, network name, and instance name. Data files installed on a clustered drive that is belonging to the same cluster group as the virtual server.
Active/Passive cluster configurations

When the NetWorker User for SQL Server program starts on the primary node, the NMSQL automatically uses the cluster virtual server as the client name for reading or writing to the NetWorker media database and client file index. If the SQL Server fails over to the secondary cluster node, opening the NetWorker User for SQL Server program on the secondary node also automatically uses the virtual server name.
Active/Active cluster configurations

If there is one SQL Server virtual server running on each physical node in the cluster, an active/active cluster configuration exists, and the following occurs:
The NMSQL automatically communicates with the virtual server running on the same physical node as the NMSQL software if no failover has occurred and each virtual server is running on a different physical node. The Select SQL Instance option from the Operation menu is disabled.
5-4
However, if one of the physical cluster nodes goes offline (for example, if a failover occurs) and both SQL Server virtual servers are then running on the same physical cluster node when the NetWorker User for SQL Server program starts, the Select SQL Instance dialog box appears. It lists the SQL Server instances. After selecting an instance from this dialog box, the NMSQL attempts to validate the instance as a NetWorker client.
How to run a scheduled backup

To schedule a backup of SQL Server data within a Microsoft cluster, use either the NetWorker Configuration Wizard or the NetWorker Administrator program. Refer to the appropriate release supplement, for instructions on using the NetWorker Configuration Wizard, to configure a scheduled backup.
Note: When using the NetWorker Configuration Wizard to configure a scheduled backup, do not use a short name alias for virtual server nodes that are not already registered on the NetWorker server with a fully qualified domain client name.
Requirements for scheduled backups

There are several requirements for performing scheduled database backups using a SQL virtual server in a cluster environment:
A NetWorker client, configured for file system backups, must be created for each physical node of the cluster. The client must be configured by using the fully qualified domain name. Configuration of the physical nodes for file system backup is required to back up nondatabase files on the cluster nodes. In the event of a disaster or a less significant system error, recovery of all or parts of the file system may be necessary. Configure the required physical node clients to the predefined save set of all. This setup must be duplicated for each physical node to insure operations continue successfully after a failover to another node.
A NetWorker client, configured for database backups, must be created for each SQL virtual server to be protected. Configure the NetWorker client to either:
5-5
Back up all databases by using MSSQL: Back up specific databases by using MSSQL: User Database. Additional databases may be added to an existing NetWorker client or a new database client may be created.
The client must be configured by using the short name for the SQL virtual server, and omitting the domain specification. If the NetWorker Configuration Wizard is used to schedule a backup of SQL Server data within a Microsoft cluster, the client running the wizard must have Modify NetWorker administrator rights. This must be set before the wizard is run. The Edit Usergroup Administrators dialog box includes these settings. Figure 5-1 on page 5-6 shows an example of the NetWorker release 7.2, Edit Usergroup Administrators dialog box.
Figure 5-1
Edit Usergroup Administrators dialog box
Figure 5-2 on page 5-7 shows the NetWorker release 7.3, Properties dialog box for administrators.
5-6
Figure 5-2
Properties dialog box for Administrators
Sample configuration for scheduled backups

Note: Figures 5-4 through 5-6 show NetWorker release 7.2 Client resource dialog boxes. Figure 5-3 on page 5-8 shows the analogous dialog box for NetWorker release 7.3. The paragraphs that follow will refer to the various tabs in reference to NetWorker release 7.3 Client resources.
5-7
Figure 5-3
NetWorker release 7.3, Client Resources Properties dialog box
A SQL virtual server in a cluster environment has the cluster name of PASS. The individual nodes are:

wash-2k.belred.legato.com cas-w2k-belred.legato.com
The SQL virtual servers are:

PASSSQL PASSSQL2
The remote user setting is for PASS_ADMIN.

Note: PASS_ADMIN is a domain account that provides sufficient rights to back up and restore databases.
Figure 5-4 on page 5-9 provides an example of an Edit Client dialog box that shows one of the NetWorker file system clients that was created for a physical node of the cluster. The fully qualified domain
5-8
name must be specified. With NetWorker release 7.3, the name field is also on the General tab (see Figure 5-3 on page 5-8). This client configuration must be duplicated for the other cluster node, cas-w2k-belred.legato.com, to ensure operations continue successfully after failover to the other cluster node.
Figure 5-4
Edit Client dialog box for Cluster Node
5-9
Figure 5-5 on page 5-10 provides an example of the Edit Client dialog box with one of the NetWorker clients was been created for each virtual server. Notice that the short client name, PASSSQL, has been specified in the Name attribute. The MSSQL: save set, for all databases can be used in the Save set attribute. Specific databases can be specified by using MSSQL:UserDatabase, where UserDatabase can be the name of any specific database. For NetWorker release 7.3, all of these fields are also on the General tab, except for the client ID, which is on the Globals (1 of 2) tab (see Figure 5-3 on page 5-8). This client is backed up when the cluster group is run. Another client is defined for PASSSQL2, the other virtual server.
Figure 5-5
Edit Client dialog box for Virtual Server
5-10
Figure 5-6 on page 5-11 provides an example of the Edit Client dialog box that shows the PASSSQL NetWorker clients Remote Access attribute. Notice that the Remote Access attribute must list both node names to provide for continuance after failover. The Remote User attribute must list the appropriate account, and the password must be listed in the Password attribute. The backup command must be included in the Backup command attribute. For NetWorker release 7.3, these fields are on the Globals (1 of 2) tab, Globals (2 of 2) tab, and the Apps & Modules tab (see Figure 5-3 on page 5-8).
Note: As shown, the remote access administrative permissions must be set up for all nodes as well as the virtual server NetWorker clients. In this example, the fully qualified domain name is used.
Figure 5-6
Edit Client dialog box
5-11
Configure scheduled backups

To configure scheduled backups by using the NetWorker Administrator program, complete the following tasks:

Task 1: Create group resources for the cluster on page 5-12 Task 2: Create Client resources for each physical node of the cluster on page 5-13 Task 3: Create Client resources for a virtual server on page 5-13 Task 4: Configure resources for snapshot backups (optional) on page 5-14 Task 5: Start the Backup on page 5-15
For detailed instructions on configuring NetWorker server resources, refer to the appropriate administrators guide. Task 1: Create group resources for the cluster For instructions on creating a Group resource, see Task 2: Configure one or more group resources on page 3-10. Modify groups so that all clients that run the NetWorker Configuration Wizard have Modify NetWorker administrator rights. This must be done before the wizard is run. Figure 5-1 on page 5-6 and Figure 5-2 on page 5-7 show examples of the NetWorker release 7.2 and 7.3 Edit Administrator List dialog boxes with Modify NetWorker administrator rights. To edit a user group: 1. For NetWorker release 7.2, in the Administrator program: a. Click the User Groups icon in the Configure tab. b. Right-click the main User Groups icon, and select Edit. 2. For NetWorker release 7.3, in the Administrator program: a. Start the NetWorker Management Console. b. Under Enterprise, select the computer to update. c. Double-click NetWorker. d. Select User Groups in the left pane, and double-click Administrators in the right-pane.
5-12
3. Edit the attributes for the group. For the Administrators group, change the Comment or the Users attribute. The Privileges attribute cannot be changed. For the Users group, all attributes can be changed. 4. Click OK. Task 2: Create Client resources for each physical node of the cluster For information about Client resource settings, see Task 3: Configure one or more Client resources on page 3-11. To create Client resources, edit the Client resource for each physical node of the cluster. In NetWorker release 7.3 Administration program, you can select Clients in the left pane, and the right-click in the right pane to create a new resource. Provide values for these attributes (on the General tab for both NetWorker release 7.2 and 7.3):
In the Name attribute of the Client resource, list the fully qualified domain name for the SQL virtual server. For example:
wash-2k.belred.legato.com
In the Save set attribute, enter all.
Task 3: Create Client resources for a virtual server To create Client resources: 1. Create the Client resource for each virtual server in the cluster. Provide values for the following attributes: a. In the Name attribute of the Client resource, list the short name for the SQL virtual server, and omit the domain specification. This should be the name of the virtual server, and not a node name. b. In the Backup command attribute on the Remote tab (for NetWorker release 7.2) or the Apps & Modules tab (for NetWorker release 7.3), enter the nsrsqlsv command with the necessary options. Command syntax for nsrsqlsv on page A-4 provides more information on the nsrsqlsv command syntax. For virtual server backups, the -a virtual_server_name command option is required.
5-13
c. On the Remote tab (for NetWorker release 7.2) or the Globals (2 of 2) tab (for NetWorker release 7.3), grant access to all physical nodes in the cluster by adding entries similar to the following in the Remote Access attribute: Remote access: RemoteUser@physicalnode_hostname Remote user: RemoteUser (Apps & Modules tab if NetWorker release 7.3) Password: ******** where: RemoteUser is the account under which the backup will be run, as shown in Figure 5-4 on page 5-9. physicalnode_hostname is the fully qualified domain name. The Remote Access attribute enables the NMSQL server to access the cluster node to authenticate the computer (on which the virtual server is running) as a NMSQL client before any backup or restore operation begins. Follow this step this for each virtual server Client resource in the cluster. d. On the Remote tab (Apps & Modules tab for NetWorker release 7.3), for the User Name and Password attributes, add the username and password, respectively, for a Windows user account that has both SQL Server administrator privileges and Windows administrator privileges. The User Name and Password attributes enable the NMSQL to back up the SQL Server virtual server. Follow this steps for each virtual server Client resource in the cluster. Task 4: Configure resources for snapshot backups (optional) To perform snapshot backups in a Microsoft cluster, NetWorker release 7.0 (or later) requires that you configure the following: 1. Add the following to the NetWorker Administrators Users Group for each physical node:
LocalSystem@physicalnode_hostname
where physicalnode_hostname is the fully qualified domain name.
5-14
2. For any physical node that is configured to use a Backup account (in the User Name and Password attributes on the Remote tab of the Client resource), add the following to the NetWorker Administrator Users Group:
User_Name@physicalnode_hostname
where: User_Name is the name of the user who has permission to backup the SQL Server clustered environment. physicalnode_hostname is the fully qualified domain name. Task 5: Start the Backup Use the NetWorker Administrator program to start the backup group manually, or wait for the next scheduled backup to occur. For instructions on starting a backup group, see one of the following:

Starting a group on Windows on page 3-18 Starting a group on UNIX on page 3-19
How to run manual backups and restores for a clustered SQL Server
A manual backup or restore operation can be run from any computer in the cluster by using one of the following:

NetWorker User for SQL Server program The nsrsqlsv and nsrsqlrc commands
Note: Manual backups and restores can be done from any computer in the cluster, regardless of whether a failover has occurred. This includes the same computer configured the same as the virtual server.
Start a backup or restore from the NetWorker User for SQL Server program
To start a manual backup or restore operation in a Microsoft cluster: 1. Start the NetWorker User for SQL Server program. The Select SQL Instance dialog box appears. 2. Select the SQL Server instance for a backup or restore operation.
5-15
3. Configure and run the backup or restore just as you would on a stand-alone server. For instructions, see the following: Performing a backup with NetWorker User for SQL Server on page 2-3 Performing a restore with NetWorker User for SQL Server on page 4-8
Start a manual backup or restore from the command prompt

To back up or restore a SQL Server virtual server, the Windows account that is used to execute the nsrsqlsv or nsrsqlrc commands must be an MSCS administrator. To determine which accounts have MSCS administrator privileges, refer to the MSCS online help. If the Windows account does not have MSCS administrator privileges, the SQL Server modules cannot communicate with MSCS and the various MSCS cluster resources, including the SQL Server virtual servers. To start a manual backup or restore:
Enter one of the following at the prompt: To back up data, enter the nsrsqlsv command. To restore data, enter the nsrsqlrc command. For example, to back up a SQL Server database, enter the following:
nsrsqlsv -a virtual_server_name -s NetWorker_server_name MSSQL:dbName
where: virtual_server_name is virtual server name when SQL Server is configured to run in an MSCS cluster. NetWorker_server_name is hostname of the NetWorker server to use for the backup. dbName is name of the SQL Server database to be backed up.
5-16
Specifying -a virtual_server_name causes the NMSQL software to do the following: Contact the SQL Server virtual server. Create save set entries under virtual_server_name in the NetWorker client index. Overview of the NMSQL commands on page A-2 provides more information on the nsrsqlsv and nsrsqlrc commands.
5-17
5-18
6
Invisible Body Tag
This chapter addresses the following Microsoft SQL Server topics:

Microsoft SQL Server recovery models ..........................................6-2 Microsoft SQL Server 2000 and 2005 named log marks ...............6-7 Transaction log maintenance............................................................6-7 SQL Server master database maintenance ................................... 6-11 Database consistency check............................................................ 6-11 Multiple Nonclustered Instances of SQL Server .........................6-13 How a restore interacts with SQL Server .....................................6-15
Note: Also, consult the most recent information from Microsoft on their SQL Server products.
6-1
Microsoft SQL Server recovery models

Microsoft SQL Servers 2000 and 2005 support three recovery models: full, bulk_logged, and simple. For SQL Server version 7.0 and earlier, similar functionality was provided by combining settings for the trunc.log.on.chkpt and select into/bulk copy database options. For further information on SQL Server version 7.0 database options, refer to Microsoft SQL Server Books Online.
Full recovery model
The full recovery model imposes the fewest constraints on the backup and restore process, but it requires the most log space of all recovery models. The NMSQL enforces the following constraints based on the version of SQL Server:
All levels of file, filegroup, and database backup are supported. SQL Server 7.0 has different constraints for filegroup and file backups at other levels. All backup data is restorable up to the most recent transaction log. Point-in-time and named log mark restores are supported for SQL Servers 2000 and 2005 only.
Bulk_Logged recovery model

The bulk_logged recovery model also imposes a few backup constraints and supports reduced log space requirements for bulk operations. The NMSQL enforces the following constraints differently, depending on which version of SQL Servers is managing the storage hierarchy.
The following backup levels are supported: For SQL Servers 2000 and 2005, all levels of file, filegroup, and database backups. For SQL Server 7.0, only level full and differential database backup types.
The following levels for backup data can be restored: All SQL Servers 2000 and 2005 backup data is restorable up to the most recent transaction log. Only SQL Server 7.0 backup data up to the last level full or differential database backup.
6-2
Point-in-time and named log mark restores are not supported for SQL Server 7.0. A point-in-time restore is not supported for SQL Servers 2000 and 2005 if the following conditions apply: If a bulk log change has occurred for the transaction log backup that corresponds to the current browse time. If bulk changes in the transaction log that contains the time or day marker. The NetWorker Module rolls forward all transactions to the end of the transaction log and leaves the database in an operational state.
Simple recovery model

The simple recovery model provides the fastest performance by minimally logging operations, thereby reducing log space. However, the simple recovery model does not support transaction log backups. The simple recovery model implements the most constraints on the backup and restore process. It provides significant control over how storage management operations impact system resources. The NMSQL enforces the following constraints depending on the version of SQL Server:
Only level full and differential database backup types are supported. Backup data is restorable up to the last level full or differential database backup. Point-in-time and named log mark restores are not supported
Specifying database recovery models

SQL Servers 2000 and 2005 allow a different recovery model to be specified for each database and for the legacy database options trunc.log.on.chkpt and select into/bulk copy. Depending on how the legacy options are set in SQL Server, they map to the new recovery models shown in Table 6-1 on page 6-4. This table describes the SQL database recovery models and how older database properties map to new properties. The transaction log operations are only available for databases with the full recovery model. All other databases only
6-3
support full database backup; incremental backups are not supported.

Table 6-1
Legacy database options Select into/bulk copy Database option True trunc.log.on.chkpt False True Simple Bulk_Logged False Simple Full
Whether or not a transaction log and the descendent filegroups and files of the database are available for backup or restore, depends on the recovery model specified in SQL Server. Based on the constraints enforced by the recovery model, the NMSQL modifies how items in the SQL Server storage hierarchy are displayed in the Backup and Restore windows. For more information about how these constraints are visually interpreted in the NetWorker User for SQL Server program, see Fake objects on page 1-20. The NMSQL enforces different backup and restore constraints, depending on which version of SQL Server is being managed. Table 6-2 on page 6-4 provides an overview of the available backup types that NMSQL supports for the recovery models available in SQL Server 2000 and 2005:
Table 6-2
Supported backup types for SQL Servers 2000 and 2005 recovery models Backup Type Database Filegroup File Full All levels All levels All levels Bulk_Logged All levels All levels All levels Simple Full and Differential None None
6-4
Table 6-3 on page 6-5 provides an overview of the various backup and restore functions that NMSQL supports for the recovery models available in each version of SQL Server.
Table 6-3
Supported backup and restore functions for SQL Server recovery models Function Transaction Log Backup Point-in-time Restore Named Log Mark Restore Full Yes Yes Yes Bulk_Logged Yes Maybea Maybe Simple No No No
a. Point-in-time restore is not supported if bulk changes are present in the transaction log that contain the time or day marker.
Microsoft SQL Server documentation provides more information about setting recovery models for SQL Server data.
Changing the recovery model for a database

SQL Server allows the recovery model for a database to be changed. However, the NMSQL does not enforce the constraints of the new recovery model after a change. Therefore, changing the recovery model might impact the current backup and restore strategies for a database. The following sections provide instructions for preparing for the recovery model change, as well as backup strategies for maintaining the consistency of the SQL Server storage hierarchy after the change. Change from full or bulk_logged to simple To change to the simple recovery model from full or bulk_logged: 1. Before changing the recovery model, perform an incremental database backup to maintain the transaction log files. 2. In SQL Server 2000 or SQL Server 2005, change the recovery model to simple. 3. After changing the recovery model, modify the backup strategy to halt execution of level incremental database backups. If the change to simple is only temporary, it is not necessary to modify the backup schedule, because the backup level is automatically
6-5
promoted to full if the recovery model (or legacy database status bit) does not support the specified level. Change from bulk_logged to full Follow these steps when changing from bulk_logged to the full recovery model: 1. In SQL Server 2000 or 2005, change the recovery model to full. 2. If performing a point-in-time restore is anticipated, perform a level incremental database backup to maintain the transaction log files. It is not necessary to modify the backup strategy. Change from simple to full or bulk_logged Because the simple recovery model has many more constraints than the full and bulk_logged models, the NMSQL might allow certain scenarios to occur. However, under normal conditions these situations would be flagged as warnings or failures by the SQL Server. The following scenarios highlight the importance of modifying the backup strategy after changing from simple to a full or bulk_logged recovery model:
Incremental backups can be created if the most recent level full or differential database backup was created when the database recovery model was simple. SQL Server enables the transaction log backup to occur, but displays a warning. Restore of a level incremental database backup cannot be performed if the level full or differential database backup that is being restored was created when the database recovery model was simple. This is consistent with SQL Server no recover behavior.
Note: The restore must be reinitiated by using the most recent level full database backup. All database transactions performed after the backup is created will be lost.
6-6
To change from simple to a full or bulk_logged recovery model: 1. In SQL Server 2000 or 2005, change the recovery model to full or bulk_logged. 2. Perform a level full or differential database backup. 3. Modify the backup strategy to include level incremental database backups.
Microsoft SQL Server 2000 and 2005 named log marks

Microsoft SQL Server 2000 and 2005 enable enhanced point-in-time restore operations by allowing named log marks to be specified during transaction creation. Database applications create named log marks when transactions are performed. The marks enable access to specific transaction points in a database transaction log backup. The NMSQL provides support for restoring to the beginning or end of a named log mark during a database restore. Restoring data by using named log marks is an improvement over point-in-time restore. The time associated with restoring to a specific transaction can be more accurately determined. When a named log mark is created in the SQL Server msdb, the log mark time is saved to the millisecond. However, the NetWorker softwares time format, which is used to specify point-in-time restore, only supports granularity to the second. If named log marks with duplicate names are created within a second of each other, the NMSQL restores to the most recently named log mark.
Transaction log maintenance

The NMSQL provides implicit and explicit methods for managing SQL Server database transaction logs:
Implicit management uses log backups to manage log space. This management can occur when: A backup schedule is implemented that includes incr (transaction log level) backups. You run the nsrsqlsv command with the -l incr option.
Microsoft SQL Server 2000 and 2005 named log marks
6-7
Explicit management specifies the nsrsqlsv command on the command line, with or without the -T option (Truncate Only) or the -G option (No Log). Both command options result in the log being truncated before the backup, and both options are compatible with level full and level 1 (differential) backups. After the backup completes, the SQL Server truncates the transaction log if the following applies: The NMSQL determines that the database does not support transaction log backups. No -T or -G option is specified.
Note: SQL Server databases that use the simple recovery model do not use transaction log backups.
How to prevent log overflow

In Windows, prevent database logs from overflowing available log space by creating an alert in the SQL Server Performance Monitor that forces an incremental backup when the databases log space reaches a certain capacity (for example, 80% full). An alert is a user-defined response to a SQL Server event. An incremental (transaction log) backup truncates the logs and clears disk space. How to create an alert for SQL Server 7.0 To create an alert for SQL Server 7.0: 1. Create a batch file (for example, sqlalert.bat) that contains the nsrsqlsv command and any appropriate command flags. For example:
installpath\nsr\bin\nsrsqlsv.exe -s NetWorker_server_name -l incr -b pool_name MSSQL:dbName PAUSE
where: installpath is the pathname where the NMSQL software is installed. The pause command is optional. NetWorker_server_name is the hostname of the NetWorker server to use for the backup.
6-8
pool_name is the backup volume to which the data is to be saved. dbName is the name of the database to backup. 2. In the Microsoft SQL Server program group, select SQL Performance Monitor. 3. From the View menu, select Alert. 4. From the Options menu, select Alert. Change the Periodic Update text box to 10 minutes or more, to enable enough time to run an incremental (transaction log) backup. 5. From the Edit menu, select Add to Alert. 6. In the Add to Alert dialog box: a. From the Objects menu, select SQL Server: Databases. b. From the Counter menu, select Percent Log Used (%). c. In the Alert If group box, click the Over button and enter 80. d. In the Run Program on Alert text box, enter the name and full path of the batch file, for example:
C:\win32app\nsr\bin\sqlalert.bat
e. In the Instance text box, select the database to monitor. f. In the Run Program on Alert group box, select Every Time. g. Click Add, then click Done. How to create an alert for SQL Server 2000 or 2005 To create an alert for SQL Server 2000 or 2005: 1. Create a batch file (for example, sqlalert.bat) that contains the nsrsqlsv command and any appropriate command flags. For example:
installpath\nsr\bin\nsrsqlsv.exe -s NetWorker_server_name -l incr -b pool_name MSSQL:db_Name PAUSE
Transaction log maintenance
6-9
where: installpath is the pathname where the NMSQL software is installed. The pause command is optional. NetWorker_server_name is the hostname of the NetWorker server to use for the backup. pool_name is the backup volume to which the data is to be saved. dbName is the name of the database to backup.
Note: Appendix A, NetWorker Module Commands provides the complete syntax of the NetWorker Module nsrsqlsv and nsrsqlrc commands, and an explanation of the command options.
2. Do one of the following: For SQL Server 2000, select Profiler in the Microsoft SQL Server program. For SQL Server 2005, select SQL Server Profiler in the Microsoft SQL Server 2005/Performance Tools program. 3. From the Tools menu, select Performance Monitor. 4. In the left pane, double-click Performance Logs and Alerts. 5. Click Alerts. Any current alerts appear in the right pane. 6. Right-click Alerts and select New Alert Settings from the shortcut menu. 7. In New Alert Settings dialog box, enter the name of the new alert in the Name text box, and then click OK. 8. Under the General tab, click Add. 9. Set the options in the Select Counters dialog box as follows: a. From the Performance Object list, select SQL Server: Databases. b. Under the Select Counter From List option, select Percent Log Used. c. Under the Select Instances From List option, select the relevant database. d. Click Add, and then click Close.
6-10
10. From the Alert When the Value Is list, select Over. 11. In the Limit text box, enter 80. 12. In the Interval text box, enter 10 (or more) to enable enough time in minutes to run an incremental (transaction log) backup. 13. From the Units list, select Minutes. 14. Under the Action tab, click Run This Program, then enter the full path of the batch file. For example:
C:\Program Files\nsr\bin\sqlalert.bat
15. Click OK.
SQL Server master database maintenance

The master database contains information about all SQL Server databases on the SQL Server host. The master database can be restored in one of the following ways:
For SQL Server 2000, by using the SQL Server Enterprise Manager to rebuild the master database. For SQL Server 2005, by using the Microsoft SQL Server Management Studio to rebuild the master database. The Microsoft SQL Server documentation provides detailed instructions on rebuilding the master database. By restoring the data for the master database and other SQL Server databases from the NetWorker server. Recovery of SQL Server on page 7-9 provides more information on restoring the master database.
Database consistency check

When running Microsoft SQL Server 7.0, use the -j option with the nsrsqlsv and nsrsqlrc commands to request that SQL Server run a comprehensive database consistency check (DBCC) before a backup or after a restore. The DBCC includes the following automatic checks:

DBCC CHECKDB database_name DBCC CHECKALLOC database_name DBCC TEXTALL database_name DBCC CHECKCATALOG database_name
SQL Server master database maintenance
6-11
For the entire DBCC to complete successfully, each of these tests must succeed.
Note: By default, the option to run a DBCC is disabled. Perform a database consistency check on page 6-12 provides information about how to trigger a DBCC before a scheduled save is initiated.
If this option is enabled and the DBCC is completed successfully, the NMSQL proceeds with a backup of the specified databases. If the DBCC does not complete successfully, the backup is terminated and the NMSQL displays a message to either:

The display to indicate the DBCC success or failure. The Group Detail dialog box in the NetWorker Administrator program.
Database consistency on a restored database can also be checked by initiating a restore from the command line on the restore host by using the -j command line option with the nsrsqlrc command. This command invokes the DBCC on the database after the restore operation finishes. You cannot check database consistency from the NetWorker User for SQL Server program. Appendix A, NetWorker Module Commands provides more information about nsrsqlsv and nsrsqlrc command options.
Perform a database consistency check

To perform a DBCC from the NetWorker Administrator program before a scheduled backup:
Using NetWorker 7.2 or earlier: a. Go to the appropriate Client resource. b. In the Edit Client or Create Client dialog box, select the Remote tab. c. In the Backup Command attribute, enter nsrsqlsv -j.
Using NetWorker 7.3 or later: a. From the Administration window, click Configuration. b. In the expanded left pane, click Clients.
6-12
c. In the right-side pane, right-click the client you want, and select Properties. d. In the Apps & Modules tab of the Properties dialog box, in the Backup Command attribute, enter nsrsqlsv -j. e. Click OK. To perform a DBCC from the Windows command prompt before a manual backup: 1. Log on to the client host with administrator privilege. 2. Enter nsrsqlsv -j at a Windows command prompt. For large databases, the running of a database consistency check can take several hours. For production systems, the -j option should be used with discretion.
Multiple Nonclustered Instances of SQL Server

Microsoft SQL Server 2000 and 2005 enable you to run multiple copies of SQL Server on a single computer. This feature is called multiple instance. The first installation of SQL Server on a computer is called the default instance. The name of the default instance is the network name for the local computer. The NMSQL provides support for backup and recovery of SQL Server 7.0, and SQL Server 2000 or 2005 default instances. Additional installations of SQL Server on a computer are called named instances. The NMSQL provides support for recovery from the default instance or named instances of SQL Server 2000 and 2005, using a copy restore, to any instance of SQL Server. This includes recovery operations when the destination server is different from the source.
Note: SQL Server 7.0 has no named instance support. Therefore, copy restore for SQL Server 7.0 applies only to/from the default instance.
Multiple Nonclustered Instances of SQL Server
6-13
Each named instance has a unique instance name in the form:

computerName\instanceName
where:

computerName is the network name of the computer. instanceName is the logical name of the named instance.
Note: When naming a SQL database or an instance, select names that are unique. Poor database name choices, for example, include the name SQL Server uses to identify itself (MSSQL:) and names of installed SQL instances you have installed.
The syntax for specifying a nonclustered instance of SQL Server at a command prompt is as follows:
MSSQL$instanceName:[dbName ...][.fgName ...][.fileName ...]
An entry of MSSQL: for the Save Set attribute in the Client resource yields a backup of all databases on the SQL Server host. For detailed guidelines on specifying the save sets for an instance, see Backup and restore command syntax for SQL Server data on page A-16. The following restrictions apply when running multiple instances:
For SQL Server 2000, Microsoft supports up to 16 named SQL Server instances to run simultaneously. For SQL Server 2005, up to 50 named instances can be run. The nsrsqlsv and nsrsqlrc commands only support specification of one instance at a time. If save sets for more than one instance are specified, the backup or restore operation fails.
Index entries for nonclustered named instances are created by using the local host on which the instance is running. To differentiate backups for the default instance and named instances, the index name has been extended to logically partition the index for a client. All running named instances are maintained in the client file index directory, excluding clustered instances and the default instance. This named instance directory is created at the end of each traditional or PowerSnap Module backup. Running nsrinfo after backups verifies the existence of this directory, for example:
%SystemDrive% nsrinfo -V -L -n mssql current SQL server instance
6-14
The sample output from the command provides information about the named instance directories TWO and THREE:
MSSQL: $/, size=252, off=400, app=mssql (14), date=1100712016 Wed Nov 17 09:20:16 2006 da_dir1-> current SQL server instance\TWO da_dir1-> current SQL server instance\THREE
How a restore interacts with SQL Server

The NetWorker Module starts and stops the SQL Server and dependent services when a restore takes place. The following sections provide details on how a NetWorker Module restore interacts with the SQL Server and dependent services:

Restoring the SQL Server master database on page 6-15 Restoring the SQL Server master database in a cluster on page 6-16 Restoring the SQL Server msdb database on page 6-16 Restoring both the SQL Server master and msdb databases on page 6-16
Restoring the SQL Server master database

When restoring the SQL Server master database, the nsrsqlrc program automatically stops and restarts the SQL Server services appropriately, as follows: 1. Before the restore begins, the NMSQL stops the SQL Server and other dependent services. 2. The NMSQL starts the SQL Server in single-user mode. 3. The NMSQL performs the restore. 4. After the restore finishes, the NMSQL waits for the SQL Server to shut down. 5. The NMSQL restarts the SQL Server in production mode.
Note: When restoring the master database, there can be timing issues related to stopping and starting of services. Manually stop all SQL Server services, except for SQL Server itself, before initiating the restore.
6-15
Restoring the SQL Server master database in a cluster

When restoring the SQL Server master database running in a cluster configuration, the nsrsqlrc program controls the SQL cluster resources, as follows: 1. Before the restore begins, the NMSQL stops the SQL Server and all dependent cluster resources, including the SQL Server Agent. 2. The NMSQL starts the SQL Server in single-user mode. 3. The NMSQL performs the restore. 4. After the restore finishes, the NMSQL waits for the SQL Server to shut down. 5. The NMSQL restarts the SQL Server and dependent cluster resources.
Restoring the SQL Server msdb database

When restoring the SQL Server msdb database, the nsrsqlrc program automatically stops and restarts the SQL Server services appropriately, as follows: 1. Before the restore begins, the NetWorker Module stops the SQL Server Agent. 2. The NetWorker Module performs the restore. 3. After the restore finishes, the NetWorker Module restarts the SQL Server Agent.
Restoring both the SQL Server master and msdb databases

When restoring the SQL Server master database along with the msdb database, the nsrsqlrc program automatically stops and restarts the SQL Server appropriately, as follows: 1. Before the restore begins, the NMSQL stops the SQL Server and other dependent services, including the SQL Server Agent.
6-16
2. The NMSQL starts the SQL Server in single-user mode. 3. The NMSQL restores the master database. 4. After the restore finishes, the NMSQL waits for the SQL Server to shut down. 5. The NetWorker Module restarts the SQL Server in production mode. 6. The NMSQL restores the msdb database. 7. The NMSQL restarts the dependent services, including the SQL Server Agent.
6-17
6-18
7
Invisible Body Tag
Disaster Recovery
This chapter provides procedures for recovering SQL Server data in the event of a disaster. This chapter includes the following sections:

Disaster recovery features.................................................................7-2 Procedures for disaster recovery .....................................................7-4
Note: The SQL Server databases must be restored in the correct order. Follow the database restore steps in Chapter 4, Restoring SQL Server Data.
Because of the many variations of system configurations, providing recovery instructions for all possible disasters is not practical. The examples that follow provide general principles and procedures for restoring data. Before beginning a SQL Server disaster recovery, review the following:

NetWorker Disaster Recovery Guide Microsoft SQL Server Books Online NetWorker Release Notes (for the appropriate system version)
Disaster Recovery
7-1
Disaster Recovery
Disaster recovery features

The NetWorker Module for Microsoft SQL (NMSQL) provides the following features for disaster recovery:
System database restore automation

Certain system databases require SQL Server service control, including the master and the msdb databases. NMSQL automates the control of these services as follows:
For the master database, the SQL Server restarts in single-user mode as required by SQL Server. For the msdb database, the SQL Agent shuts down to close connections to the msdb database. How a restore interacts with SQL Server on page 6-15 provides more information on how the NMSQL controls SQL Server services.
Note: The NMSQL does not support snapshot backups or recoveries of the SQL Server master database. Use a traditional backup to restore the master system database during a disaster recovery.
Database restore order

When restoring a complete backup of all databases, or when restoring certain system databases, the restore must occur in a specific order. When system databases are present in the restore list, the NMSQL ensures that the restore order follows SQL Server procedures as follows: 1. The master database is always restored first. This ensures that metadata present in the master database is correct for all subsequent restored databases. 2. The msdb database is always restored after the master database and before all other databases. This ensures that scheduling and other system data present in the msdb database are correct for all subsequent restored databases.
7-2
Disaster Recovery
3. The model database is always restored after master and msdb databases, and before all other databases. This ensures that the database configuration is correct for all subsequent restored databases.
SQL Server startup complete detection

When the SQL Server starts, it launches a database startup process. The SQL Server enables user connections while the startup process is running. However, if the startup process is interrupted by a database RESTORE query, any database that has not yet started is marked as suspect. When the interruption occurs, subsequent restores of the msdb database fails. Subsequent restores of any user database might also fail unless the Overwrite the Existing Database attribute is specified in the Properties dialog box. The NMSQL waits for SQL Server to complete the database startup process for all databases before starting a restore. The NMSQL waits for the restore of the master database to complete. This wait process enables a proper restore of the msdb database and user databases following a SQL Server startup.
Note: Always use the -f option with the nsrsqlrc command for restore operations that follow a SQL Server startup.
Overwriting capability
Use the Files tab in the Properties dialog box of the NetWorker User for SQL Server program to specify overwriting all databases during a restore. For further details, see the Properties dialog box in Figure 4-25 on page 4-43.
Note: The NMSQL does not handle all dependent services. When restoring application services, such Metadata or Replication services, as well as the databases on which these services depend, they must be manually shut down.
The NMSQL does not ensure that all connections to a database are closed prior to restoring a database as required by SQL Server. Such open connections must be manually terminated. The Microsoft SQL Server Books Online provides more information.
Disaster recovery features
7-3
Disaster Recovery
Procedures for disaster recovery

Use the following instructions for disaster recovery if NetWorker Module binary files or SQL Server binary files are damaged or lost:

When not to reinstall SQL Server on page 7-4 Recovery of a damaged primary disk of the SQL Server on page 7-5 Recovery of a damaged primary disk of the SQL Server on page 7-5 Recovery of a damaged binary disk of the SQL Server on page 7-6 Recovery of SQL Server and NetWorker server on page 7-7 Recovery of SQL Server without reinstalling SQL Server (shortcut) on page 7-8 Recovery of SQL Server on page 7-9
The instructions explain how to recover from a disaster on a NMSQL client host for the following configurations:

Windows 2000 with SQL Server 7.0 Windows 2000 or Windows Server 2003 with SQL Server 2000 or 2005
For more information about using NetWorker software for disaster recovery, refer to the NetWorker Disaster Recovery Guide.
When not to reinstall SQL Server

During a disaster recovery of SQL Server, do not reinstall SQL Server if the following applies:

A recent NetWorker backup of the ALL save set exists. SQL Server was installed prior to the creation of the ALL backup version. SQL backups for all of the SQL Server databases exist.
7-4
Disaster Recovery
If SQL Server is already installed when the ALL backup is performed, critical state information for SQL Server is backed up as part of the ALL save set. When the ALL save set contains this information, and backups for all of the SQL Server databases exist, the SQL Server can be reinstated by only restoring the ALL save set and the SQL databases.
Recovery of a damaged primary disk of the SQL Server

If the primary disk with critical SQL Server data is damaged, do the following: 1. Shut down SQL Server. 2. Run the Rebuild Master utility, rebuildm, located in the SQL \Binn directory. The Rebuild Master utility requires SQL system database files in the Data directory of the SQL Server installation CD-ROM or shared network drive. For SQL Server 2000 or 2005: Copy these files from the installation CD-ROM to a temporary location. Remove the read-only attributes. Direct the Rebuild Master utility to use the temporary location as the source directory for data files. 3. Restart SQL Server. 4. Use the replace option to restore the SQL Server master database and the msdb database. 5. Restore the other SQL application databases.
Note: To run steps 4 and 5 as a single operation from the command line, enter the following: nsrsqlrc -s... -f MSSQL: To use the NetWorker User for SQL Server to complete steps 4 and 5, see Use the NetWorker User for SQL Server Program to complete a disaster recovery on page 7-10. If running SQL Server 7.0, the model and the distribution databases might need to be restored.
7-5
Disaster Recovery
Recovery of a damaged binary disk of the SQL Server

If the disk with the SQL Server binaries is damaged, do the following: 1. Shut down SQL Server. 2. Restore the ALL save set. For more information about the ALL save set, see When not to reinstall SQL Server on page 7-4. 3. Restart the computer. 4. Shut down SQL Server. 5. Run the Rebuild Master utility, rebuildm, located in the SQL \Binn directory. The Rebuild Master utility requires SQL system database files in the Data directory of the SQL Server installation CD-ROM or shared network drive. For SQL Server 2000 or 2005: Copy these files from the installation CD-ROM to a temporary location Remove the read-only attributes. Direct the Rebuild Master utility to use the temporary location as the source directory for data files. 6. Restart SQL Server. 7. Use the replace option to restore the SQL Server master database and the msdb database. 8. Restore the other SQL application databases.
7-6
Disaster Recovery
Recovery of SQL Server and NetWorker server

If the disks with the SQL Server binaries and the NetWorker online indexes (the nsr file system) or the SQL Server data are damaged, then perform the following steps: 1. Restore the NetWorker binaries and online indexes. The NetWorker Disaster Recovery Guide provides details. 2. Reinstall the NetWorker software. If the computer that failed was the NetWorker server, reinstall the NetWorker server software. If the computer that failed was a NetWorker client, reinstall the NetWorker client software. 3. Restore the ALL save set. When not to reinstall SQL Server on page 7-4 provides more information about the ALL save set. 4. Restart the computer. 5. Shut down SQL Server services if they are running. The Rebuild Master utility requires SQL system database files in the Data directory of the SQL Server installation CD-ROM or shared network drive. For SQL Server 2000 or 2005: Copy these files from the installation CD-ROM to a temporary location. Remove the read-only attributes. Direct the Rebuild Master utility to use the temporary location as the source directory for data files. 6. Run the Rebuild Master utility, rebuildm, in the SQL \Binn directory. 7. Restart SQL Server. 8. Use the replace option to restore the SQL Server master database and the msdb database.
7-7
Disaster Recovery
9. Restore the other SQL application databases.

Recovery of SQL Server without reinstalling SQL Server (shortcut)

Use this procedure to recover the SQL Server when the server does not need to be reinstalled. When not to reinstall SQL Server on page 7-4 provides more detail. To perform a disaster recovery of the SQL Server without having to reinstall it: 1. Reinstall the operating system. 2. Reinstall the NetWorker software. If the computer that failed was the NetWorker server, reinstall the NetWorker server software. If the computer that failed was a NetWorker client, reinstall the NetWorker client software. 3. Restore the ALL save set. 4. Restart the computer. 5. Shut down SQL Server. The Rebuild Master utility requires SQL system database files in the Data directory of the SQL Server installation CD-ROM or shared network drive. For SQL Server 2000 or 2005: Copy these files from the installation CD-ROM to a temporary location. Remove the read-only attributes.
7-8
Disaster Recovery
Direct the Rebuild Master utility to use the temporary location as the source directory for data files. 6. Run the Rebuild Master utility, rebuildm, in the SQL \Binn directory. 7. Restart SQL Server. 8. Use the replace option to restore the SQL Server master database and the msdb database. 9. Restore the other SQL application databases. The SQL Server should now be restored to the most recent backup.
Recovery of SQL Server

To recover the SQL Server: 1. Reinstall the operating system. 2. Reinstall the NetWorker software. If the computer that failed was the NetWorker server, reinstall the NetWorker server software. If the computer that failed was a NetWorker client, reinstall the NetWorker client software. 3. Recover or reinstall the SQL Server software. To determine if you need to reinstall the SQL Server software, see When not to reinstall SQL Server on page 7-4. 4. Start SQL Server. 5. Use the replace option to restore the SQL Server master database and the msdb database.
7-9
Disaster Recovery
6. Restore the other SQL application databases.

Use the NetWorker User for SQL Server Program to complete a disaster recovery
To complete the disaster recovery by using the NetWorker User for SQL Server program: 1. Click Restore. 2. Select Normal, and then click Continue. The Restore (Normal) window appears. 3. Right-click the SQL Server root and select Mark All Databases. 4. From the Options menu, select Restore Options. The Restore Options dialog box appears. 5. Click Automatically Overwrite Any Database Marked for Restore, and then click OK. 6. Click Start.
7-10
Invisible Body Tag
A
This appendix provides syntax, notation, and related information for NetWorker Module for Microsoft SQL Server (NMSQL) commands and command options. Command-line examples are provided.
Note: The command-line syntax examples provided in the following sections are shown in a multiple-line format in order to improve the readability. When actually entering a command at the command prompt, the command, command options, and specified data must all be entered on a single line.
This appendix includes the following main sections:

Overview of the NMSQL commands.............................................A-2 Using the nsrsqlsv command ..........................................................A-3 Using the nsrsqlrc command...........................................................A-8 Using the nwmssql command.......................................................A-15 Backup and restore command syntax for SQL Server data ......A-16
A-1
Overview of the NMSQL commands

NMSQL commands perform the following functions:
nsrsqlsv backs up the specified SQL Server data objects.Using the nsrsqlsv command on page A-3 provides more information. nsrsqlrc restores the specified SQL Server data objects. Using the nsrsqlrc command on page A-8 provides more information. nwmssql invokes the client graphical user interface. Using the nwmssql command on page A-15 provides more information.
Descriptions, command options, and usage scenarios for each command are defined in the following sections. However, when initiating the commands, keep in mind these notes regarding syntax:
Case is very important when specifying command-line flags. Each command option is either lowercase or uppercase and, frequently, both the cases of a letter are included in the set of command options. For example, -c specifies the NetWorker client name, while -C specifies compression of the backup data. Depending on the command option, the space separator between an option and its corresponding argument can be optional, required, or not allowed. For example, the following expressions are both valid:
-l backup_level -lbackup_level
While the following expression is invalid because a space is not allowed between the + argument and log_mark_name:
-M + log_mark_name
Brackets ([ ]) are used to denote the optional portions of a command (for example, command options and corresponding arguments, if any). When initiating an actual backup or restore operation, do not include brackets. Data items must follow: All other command options. Parameters on the command line.
An application log is generated for both nsrsqlsv.exe and nsrsqlrc.exe programs. The logs are written into the nsr\applogs folder on the SQL Server host. The logs are cumulative and are
A-2
appended each time the nsrsqlsv.exe or nsrsqlrc.exe program runs. The log space management is rudimentary. The log is truncated when disk space is exhausted. Error logs for backup and recovery on page 1-25 provides more information about application logs.
Using the nsrsqlsv command

The nsrsqlsv command is used to back up SQL Server data objects, which consist of files, filegroups, and databases. To initiate a backup operation, specify nsrsqlsv and its command options for the Backup Command attribute in the Client resource or from a Windows command prompt.
Using NetWorker 7.2 and earlier

To modify the Backup Command attribute:
On Windows: 1. In the NetWorker Administrator program, go to the Configure tab and click Manage Clients. 2. Right-click the client you want and select Edit. The Edit Client dialog box appears. 3. In the Backup Command attribute on the Remote tab, enter nsrsqlsv and appropriate options. 4. Click OK.
On UNIX: 1. In the NetWorker Administrator program, select Client Setup from the Clients menu. 2. Select the client you want from the Clients list. 3. In the Backup Command attribute on the Remote tab, enter nsrsqlsv and appropriate options. 4. Click Apply.
Note: The -b and -l command options are valid only for manual backups initiated from a Windows command prompt on a client host. Do not use either of these options when initiating a scheduled save in the NetWorker Administrator program.
A-3
Using NetWorker 7.3 and later

To modify the Backup Command attribute: 1. From the Administration window, click Configuration. 2. In the expanded left pane, click Clients. 3. In the right-side pane, right-click the client you want, and select Properties. 4. In the Apps & Modules tab of the Properties dialog box, enter nsrsqlsv and any needed command options in the Backup Command field. 5. Click OK.
Note: The -b and -l command options are valid only for manual backups initiated from a Windows command prompt on a client host. Do not use either of these options when initiating a scheduled save in the NetWorker Administrator program.
Command syntax for nsrsqlsv

The nsrsqlsv command syntax is:
nsrsqlsv [-CGjqRTvkuHZ] [-a virtual_server_name] [-b pool_name] [-c client_name] [-g group ] [-h dbName] [-I input_file] [-l backup_level] [-N name] [-s NetWorker_server_name] [-S stripes_count] [[-U username] [-P password] [MSSQL:dbName dbName.fgName dbName.fgName.filename ...]
A-4
Command options for nsrsqlsv

Command options are as follows:
Table A-1
Command options for nsrsqlsv Command options -a
Descriptions Specifies the virtual server name when SQL Server is configured to run in an MSCS cluster. For more information, see Chapter 5, Backup and Recovery for Microsoft Cluster Server. Assigns a destination pool for a save set. Specification of -b pool_name overrides all other pool-selection criteria either provided by the NetWorker software, or specified in the NetWorker User for SQL Server Backup Options dialog box. The pool must be created with a corresponding label template before running a command that includes the -b option. Specifies the NetWorker client name for which the SQL Server data is to be backed up. Specifies compression of the backup data before the NetWorker client sends the data to the NetWorker server. Specifies the save group. The NetWorker server and the savegrp command use the group parameter to select the media pool. Specifies a NO_LOG transaction log backup before backing up the database. This command option is valid only for level full backups. Is used to exclude a database from the backup. For example:
nsrsqlsv -s bv-aditya.belred.legato.com -h master -h model MSSQL:
-b
-c -C -g -G -h
-H -I
Uppercase -H uses the NORECOVERY option when backing up transaction logs. It leaves the database in Restoring state. Specifies a text file that lists multiple SQL Server save sets, for example: MSSQL$SQL2000:database_1 MSSQL$SQL2000:database_2 ... MSSQL$SQL2000:database_100 The -I option may also be specified with the nsrsqlsv command for the Backup Command attribute in the NetWorker Client resource. Performs a database consistency check before initiating the backup. For more information about this option, see Database consistency check on page 6-11. Perform checksum before writing to media.
-j
-k
A-5
Table A-1
Command options for nsrsqlsv (continued) Command options -l
Descriptions Specifies the backup level. Valid values are as follows: Full Level 1 (equivalent to specifying diff in BusinesSuite Module 2.0 and 2.0.1) Incr (equivalent to specifying xlog in BusinesSuite Module 2.0 and 2.0.1). For more information about which backup levels are supported for various SQL Server data objects, see Combining data objects to create backup levels on page 3-6. Note: The -l option is valid only for manual backups initiated from a Windows command prompt on a client host. For scheduled backups, set the backup level in the Set Level dialog box of the Schedule resource in the NetWorker Administrator program. Do not use the -l option when initiating a backup in the NetWorker Administrator program from the Backup Command attribute of the Edit Client dialog box (under the Remote Tab). Sample outputs for each of the three different levels of backup follow: Full backup of the database (my_database) to a specified NetWorker server: nsrsqlsv -s NetWorker_server_name my_database nsrsqlsv: Backing up my_database... nsrsqlsv: BACKUP database my_database TO virtual_device='BSMSQL' WITH stats nsrsqlsv: my_database: level=full, 33 MB 00:00:05 1 file(s) Differential backup: nsrsqlsv -s NetWorker_server_name -l diff my_database nsrsqlsv: Backing up my_database... nsrsqlsv: BACKUP database my_database TO virtual_device='BSMSQL' WITH differential, stats nsrsqlsv: my_database level=diff, 719 KB 00:00:05 1 file(s) guidelines on formatting these parameters, see Backup and restore command syntax for SQL Server data on page A-16.
A-6
Table A-1
Command options for nsrsqlsv (continued) Command options
Descriptions Incremental backup: nsrsqlsv -s NetWorker_server_name -l incr my_database nsrsqlsv: Backing up my_database... nsrsqlsv: BACKUP log my_database TO virtual_device =BSMSQL nsrsqlsv: my_database level=incr, 61 MB 00:00:05 1 file(s) At least one SQL Server data item (file, filegroup, or database) must be specified, and the data items and list of data objects must follow all other command options and parameters on the command line.
-N -P
Specifies the symbolic name of the save set. By default, the most common prefix of the path arguments is used as the save set name. Specifies the Microsoft SQL Server user password. When the -U username command option is specified, the password command option must also be provided, as follows: nsrsqlsv -s NetWorker_server_name -U username -P password MSSQL: Use the SQL Server username and password to log onto SQL Server by using SQL Server integrated security. Displays nsrsqlsv messages in quiet mode; only summary information and error messages are displayed. Uses the NO_TRUNCATE option when backing up transaction logs. Specifies the NetWorker server to use for the backup operation. Backs up the specified data items using n stripes. Note: To use backup and restore striping successfully, see the striping configuration described in Appendix C, Striped Backup and Recovery.
-q -R -s -S
-T -u
Performs a TRUNCATE_ONLY transaction log backup before backing up the database; valid for full backups only. Continue the backup even in the event of a checksum error.
A-7
Table A-1
Command options for nsrsqlsv (continued) Command options -U
Descriptions Specifies the Microsoft SQL Server username. When this command option is specified, the -P password command option must also be provided, as follows: nsrsqlsv -s NetWorker_server_name -U username -P password MSSQL: Use the SQL Server username and password to log onto SQL Server using SQL Server integrated security. Displays nsrsqlsv messages in verbose mode, providing detailed information about the progress of the backup operation. Applies to the backup of up databases for SQL Server 2005, and is usually used in the online restore scenario from the command line. The -Z option specifies that the incremental (transaction log) backup after restore is not promoted to full backup. Without the -Z option, the backup is promoted to full.
-v -Z
Using the nsrsqlrc command

The nsrsqlrc command restores specified SQL Server data (files, filegroups, and databases) from the NetWorker system. To initiate a restore operation, specify nsrsqlrc and any of its command options at a Windows command prompt.
Command syntax for nsrsqlrc

The nsrsqlrc command syntax is as follows:
nsrsqlrc [-fjqVku] [-a virtual_server_name] [-c client_name] {[-d MSSQL:destination_dbName] [-C file=path,file2=path2,...]} [-M [+|-]log_mark_name] [-R fgName1,fgName2,...] [-t date] [-s NetWorker_server_name] [-S normal|standby:undo_file|norecover] [[-U username] [-P password]] [MSSQL: dbName dbName.fgName dbName.fgName.filename ...]
A-8
Command options for nsrsqlrc

The command options are as follows
Table A-2
Command options for nsrqlrc Command options -a
Description Specifies the virtual server name when SQL Server is configured to run as a clustered service in an MSCS cluster. Chapter 5, Backup and Recovery for Microsoft Cluster Server provides more information. Specifies the NetWorker client name from which the SQL Server data is to be restored. Copies the database being restored to either the same SQL Server or a different SQL Server. It can be used for normal, copy, and partial restores. Use the relocation list to specify new locations for restored database files. The relocation list is composed of pairs of logical database filenames and fully qualified domain database filename relocation paths. Specify the relocation list only when restoring a database. Each filename and relocation path is joined by the equal sign (=), and pairs are delimited in the list by commas. The syntax is: ["][] file[]=[]path[],[] file2[]=[]path2[],...["] For example, to copy a database named Project from a client host named ClientHost1 to a different location on ClientHost1, specify the relocation list for the database files, but do not include the client host name command option: nsrsqlrc -s NetWorker_server_name -d MSSQL:CopyOfProject -C Project_Data=C:\Relocation\Path\Project_Data.mdF, Project_Data2=C:\Relocation\Path\Project_Data2.ndF,..., Project_Log=C:\Relocation\Path\Project_Log.ldF MSSQL:Project The relocation list may be enclosed by double quotes to allow for spaces in the relocation elements and pathnames. A logical filename or relocation path may be enclosed by single quotes to also allow for spaces. If a filename or path contains a single quote, precede the single quote with another single quote to prevent the NMSQL software from parsing the single quote as a delimiter, for example: nsrsqlrc -s NetWorker_server_name -d MSSQL:CopyOfProject -C"File=C:\Relocate Dir\Path\, =C:\Relocate Dir\Path\,..., =C:\Relocate Dir\Path\" MSSQL:Project
-c -C
A-9
Table A-2
Command options for nsrqlrc (continued) Command options
Description When no relocation list is specified, the NMSQL reads the source database filenames and location from the client index metadata or the backup. This information is used to generate a default list by relocating all files to the default SQL data path for the target SQL Server. The filenames are guaranteed to be unique, but sufficient disk space is not ensured.
-d
Performs a copy operation (Microsoft SQL Server 7.0 and higher). The copy operation restores SQL Server data from a client host to another database name on the same client host. The syntax is: nsrsqlrc -s NetWorker_server_name -C client_name -d MSSQL:destination_dbname MSSQL:source_dbname where: destination_dbName is the name of the SQL database to which the source database is to be restored. source_dbName is the name of the SQL database to restore. Note: When -C, -M, -R, or -d are used, the list of data items can include only one database. The list of data items must follow all other command options and parameters on the command line. Backup and restore command syntax for SQL Server data on page A-16 provides detailed guidelines on formatting these parameters.
-f
Performs a restore operation by replacing the target database with the source. This option restores a source database to an existing, incompatible database of the same name on the target host. This option is also used to restore damaged files. Performs a database consistency check between the SQL Server data backup and the restored SQL Server data. Database consistency check on page 6-11 provides more information about this option.
-j
A-10
Table A-2
Command options for nsrqlrc (continued) Command options -M
Description Performs a SQL Server data restore of the named transaction specified in log_mark_name (Microsoft SQL Server 2000 and 2005 only). How the mark name is prefixed, determines how the data will be restored: When the mark name is prefixed with a plus sign (+), the data is restored to and includes the named transaction. When the mark name is prefixed with a minus sign (-), the data is restored up to the named transaction. The log_mark_name should immediately follow the plus or minus symbol. The use of a space separator is not allowed. The default is the plus sign. For example, to restore the SQL data to and include the named transaction transaction_mark, enter the following command: nsrsqlrc -s NetWorker_server_name -M +transaction_mark MSSQL:dbName To restore the SQL data only to the named transaction transaction_mark, enter the following command: nsrsqlrc -s NetWorker_server_name -M -transaction_mark MSSQL:dbName Only one SQL Server database may be specified, and the database must follow all other command options and parameters on the command line.
-P
Specifies the Microsoft SQL Server user password. When the -U username command option is specified, the password command option must also be provided, as follows: nsrsqlrc -s NetWorker_server_name -U username -P password MSSQL: Use the SQL Server username and password to log onto SQL Server using SQL Server integrated security.
-q
Displays nsrsqlrc messages in quiet mode, which provides minimal information about the progress of the restore operation, including error messages.
A-11
Table A-2
Command options for nsrqlrc (continued) Command options -R
Description Performs a partial database restore (when using Microsoft SQL Server 2000) or a piecemeal restore (when using SQL Server 2005) of the specified filegroups. This command option is not available with other versions of SQL Server. The partial database restore operation restores specific filegroups from a single full SQL Server database backup. Supply the filegroups to the -R command option in a list, with items separated by commas. The -C parameter may be used, and should specify all files for the database. The -d parameter is also required: nsrsqlrc -s NetWorker_server_name -R ["][]fgName[],[]fgName2[],[]...[]["] -C Project_Data=C:\Relocation\Path\Project_Data. mdF, Project_Data2=C:\Relocation\Path\Project_Data 2.ndF,..., Project_Log=C:\Relocation\Path\Project_Log.ld F MSSQL:Project -d MSSQL:PartOfProject MSSQL:Project where: Project is the name of the SQL database to restore. PartOfProject is the name of the SQL database to which the source database is to be restored fgName,... are the names of the filegroups to restore. To allow spaces: Between the filegroup names for the -R option, enclose the list of filegroup names within double quotes. Within filegroup names, enclose each filegroup name within single quotes. Note: If a filegroup name contains a single quote, precede the single quote with another single quote to prevent the NetWorker software from parsing the single quote as a delimiter. For example, to accomodate for the space in Filegroup A, the space after the first comma, and the single quote in Filegroup A, use the following syntax: -R "Filegroup A, Filegroup A" When an empty relocation list is supplied, use the following syntax: -R ""
A-12
Table A-2
Command options for nsrqlrc (continued) Command options
Description The NMSQL restores only the primary filegroup. When -C, -M, -R, or -d are used, the list of data objects can include only one database. The list of data objects must follow all other command options and parameters on the command line. Backup and restore command syntax for SQL Server data on page A-16 provides detailed guidelines on formatting these parameters.
-s -S
Specifies the NetWorker server to use for the restore operation. Uppercase -S performs the restore operation according to the specified database restore mode of normal, standby, or no recovery (Microsoft SQL Server 2000 and 2005 only). The syntax is: nsrsqlrc -s NetWorker_server_name -d destination_dbName -S normal | "standby:undo_file" | norecover MSSQL:source_dbName where: destination_dbName is the name of the SQL database to which the source database is to be restored. source_dbName is the name of the SQL database to restore. The restore modes are as follows: The normal restore mode restores the database in normal, operational mode. The standby restore mode activates the SQL Server STANDBY option, which forces the database to be in a read-only state between transaction log restore operations. The no-recovery restore mode activates the SQL Server NORECOVER option, which places the database in an unloadable state after the restore, but still able to process additional transaction log restore operations. For example, to restore a database named Project in normal, operational mode to a new database named NewProjectName, enter the following command: nsrsqlrc -s NetWorker_server_name -S normal -d MSSQL:NewProjectName MSSQL:Project To restore the database in standby mode, the standby parameter must be immediately followed by a colon, and the standby undo file location and filename must be specified. If a filename or location contains a space, enclose the filename or location within double quotes, for example: nsrsqlsv -s NetWorker_server_name -S "standby:C:\temp\undo filename" -d MSSQL:NewProjectName MSSQL:Project
A-13
Table A-2
Command options for nsrqlrc (continued) Command options -t
Description Restores SQL Server data as of the specified date. When the date of a backup version occurs before or is equivalent to the date, the backup version is restored. Follow the nsr_getdate command syntax guidelines when formatting the date. To avoid adversely affecting the database, do not use the -t option if restoring a file or filegroup. Specifies the Microsoft SQL Server username. When the username command option is specified, the -P password command option must also be provided, for example: nsrsqlrc -s NetWorker_server_name -U username -P password MSSQL: Use the SQL Server username and password to log onto SQL Server by using SQL Server integrated security. Verifies the SQL Server database selected for the restore (Microsoft SQL Server 2000 and 2005 only). The -V command option verifies only that the selected database backup is suitable for restoring, the backup is not restored. The syntax is: nsrsqlrc -s NetWorker_server_name -V MSSQL:dbName Perform checksum before reading from media. Continue the restore even in the event of a checksum error.
-U
-V
-k -u
Sample restore command lines

In an active mirror session, the user interface prevents a piecemeal restore of the principal database to a different However, a piecemeal restore can be performed from the command line. Sample command line for a piecemeal restore of the primary filegroup (MDF & LDF) and filegroup "a" (NDF) of AcmeBank to the new database AcmeOnline:
nsrsqlrc -s "bv-v-cgd2.belred.legato.com" -c "bv-v-cgd2.belred.legato.com" -$ "MSSQL$THREE:" -R "'PRIMARY','a'" -d "MSSQL$THREE:AcmeOnline" -C "'AcmeBank'='E:\Data\AcmeOnline.mdf', 'AcmeBank_log'='E:\Data\AcmeOnline_log.ldf', 'AcmeBank1'='E:\Data\AcmeOnline1.ndf'" -t "Wed Sep 14 13:31:46 2005"
A-14
"MSSQL$THREE:AcmeBank" Note: The AcmeOnline database name and file locations are different from AcmeBank.
Using the nwmssql command

The nwmssql command invokes the NetWorker User for SQL Server program, the client graphical user interface. To run the NetWorker User for SQL Server program from the Windows Start menu, select Programs>EMC NetWorker>NetWorker User for SQL Server.
Note: The EMC NetWorker group referred to as NetWorker Group in NetWorker releases earlier than 7.0.
To create a desktop shortcut, go to the <install_path>\nsr\bin directory and drag the nwmssql.exe file to your desktop while pressing the [Ctrl] key.
Command options for nwmssql

The nwmssql command options are as follows:
nwmssql [-t] -s NetWorker_server_name
Table A-3
Command options for nwmssql Command options for nwmssql -s -t
Description Specifies the NetWorker server to use. Diagnoses the current backup or restore command issued by the NMSQL. The NetWorker User for SQL Server program displays the full backup or restore command in the operation status window, but does not execute the backup or restore operation.
NetWorker User for SQL Server program on page 1-20 provides more information on the NetWorker User for SQL Server program.
Using the nwmssql command
A-15
Backup and restore command syntax for SQL Server data

With the standard NetWorker backup and restore commands (nsrsqlsv and nsrsqlrc), use the additional command syntax shown in Table A-4 on page A-16 to back up or restore SQL Server data. Enter the NetWorker commands with the SQL Server data syntax for either scheduled or manual backups as follows:
Scheduled backup In the NetWorker Administrator program, enter the command in the Backup Command attribute of the Create Client or Edit Client dialog box.
Manual backup Enter the command at the Windows command prompt on the NetWorker server. At least one SQL Server data item (file, filegroup, or database) must be specified for a manual backup or restore.
You can specify more than one data object and combine different types of data. SQL data objects must be specified by using the syntax shown in Table A-4 on page A-16.
Table A-4
Command syntax for SQL Server data SQL Server data All databases in the SQL Server storage hierarchy (optional) Syntax for SQL Server data objects MSSQL: Entering only MSSQL: always yields a backup of all databases on the SQL Server host. A snapshot backup fails if more than one database, or MSSQL: is specified for the Save Set attribute. For snapshot backups, list only one database for the Save Set attribute. MSSQL:dbName or [MSSQL:dbName MSSQL:dbName2 ...] MSSQL:dbName. or [MSSQL:dbName. MSSQL:dbName2 ...]
Specified databases
All filegroups in specified databases
A-16
Table A-4
Command syntax for SQL Server data (continued) SQL Server data Specified filegroups in specified database Syntax for SQL Server data objects MSSQL:dbName.fgName or [MSSQL:dbName.fgName MSSQL:dbName.fgName2 MSSQL:dbName2.fgName MSSQL:dbName2.fgName2 ...] MSSQL:dbName.fgName.filename or [MSSQL:dbName.fgName.filename MSSQL:dbName.fgName2.filename MSSQL:dbName2.fgName.filename MSSQL:dbName2.fgName2.filename ...]
Specified files in specified filegroups in specified databases
Specifying MSSQL before each data object name is optional and does not affect the expression or the resulting operation. However, when MSSQL is specified, it must be followed by a colon (:). For example, the following two commands are equivalent:
nsrsqlsv -s NetWorker_server_name dbName.fgName nsrsqlsv -s NetWorker_server_name MSSQL:dbName.fgName
In a nonclustered, named instance configuration, MSSQL$ is required, followed by the instance name and a colon. For example:
nsrsqlsv -s NetWorker_server_name MSSQL$instanceName:dbName.fgName
Syntax for a named instance configuration for SQL 2000 and 2005
When the configuration contains nonclustered named instances of SQL Server, the name of the instance should be specified before the data, as follows:
MSSQL$instanceName:[dbName ...] [.fgName ...] [.fileName ...]
A-17
For example, to back up all of the databases for instanceOne, enter the following:
nsrsqlsv -s NetWorker_server_name MSSQL$instanceOne:
To restore several filegroups for instanceTwo, specify:

nsrsqlrc -s NetWorker_server_name MSSQL$instanceTwo:dbName.fgName MSSQL$instanceTwo:dbName.fgName2
Instead of using clustered named instances in this syntax, use clustered instance SQL Server virtual server names with -a or -c option. For example:
nsrsqlsv -s NetWorker_server_name -a SQL_virtual_server_DNS_name MSSQL:nsrsqlsv -s NetWorker_server_name -c SQL_virtual_server_DNS_name MSSQL:
where:

NetWorker_server_name is the hostname of the NetWorker server. SQL_virtual_server_DNS_name is the Domain Name System (DNS) name for the SQL Server virtual server. A Client resource should be created under this name.
For scheduled saves of a SQL Server virtual server client, it is not necessary to specify -a or -c option with the SQL Server virtual server name. The savegrp process automatically specifies the virtual server name to the nsrsqlsv process by using the -m option.
Note: The nsrsqlsv and nsrsqlrc commands only support specification of a single instance. If save sets for more than one instance are specified, the backup fails. The nsrsqlrc command supports mixing of instances for a copy restore operation. Multiple Nonclustered Instances of SQL Server on page 6-13 provides more information about running multiple instances of SQL Server.
A-18
Syntax for names containing a period, backslash, or colon

The NMSQL provides command line syntax that enables you to back up and restore filenames, filegroups, and databases containing a period (.), backslash (\), or colon (:). By entering a backslash before the period or backslash, the nsrsqlsv and nsrsqlrc commands interpret the period or backslash as a literal character. Tables A-5 through A-8 show the syntax for filenames, filegroups, and databases containing a period, backslash, colon, or any combination of the three. The following notes apply to the information in the tables:
The syntax shown in the right column applies to both the nsrsqlsv or nsrsqlrc commands. The notation MSSQL: is optional only for the nsrsqlsv command. A single period (.) continues to delimit SQL identifiers. The syntax also applies to named instances. The backslash period (\.) character sequence replaces each literal period in the SQL identifier. The double backslash (\\) character sequence replaces each literal backslash in the SQL identifier.
Table A-5
Command syntax for names containing a period Name visible from SQL utilities SQL database named MyDatabase.COM. SQL filegroup named MyFileGroup.2 for the SQL database named MyDatabase.COM. SQL file named MyFile.2, which is a member of the SQL filegroup named MyFileGroup.2 for the SQL database named MyDatabase.COM.
Equivalent command-line syntax

MyDatabase\.COM MSSQL:MyDatabase\.COM MSSQL$MyInstance:MyDatabase\.COM MyDatabase\.COM.MyFileGroup\.2 MSSQL:MyDatabase\.COM.MyFileGroup\.2 MSSQL$MyInstance:MyDatabase\.COM.MyFileGrou p\.2 MyDatabase\.COM.MyFileGroup\.2.MyFile\.2 MSSQL:MyDatabase\.COM.MyFileGroup\.2.MyFile \.2 MSSQL$MyInstance:MyDatabase\.COM.MyFileGrou p\.2.MyFile\.2
A-19
Table A-6
Command syntax for names containing a backslash Name visible from SQL utilities The SQL database named MyDatabase\COM. The SQL filegroup named MyFileGroup\2 for the SQL database named MyDatabase\COM. The SQL file named MyFile\2, which is a member of the SQL filegroup named MyFileGroup\2 for the SQL database named MyDatabase\COM. The SQL database named MyDatabase\COM.

MyDatabase\\COM MSSQL:MyDatabase\\COM MSSQL$MyInstance:MyDatabase\\COM MyDatabase\\COM.MyFileGroup\\2 MSSQL:MyDatabase\\COM.MyFileGroup\\2 MSSQL$MyInstance:MyDatabase\\COM.MyFileGroup\ \2
MyDatabase\\COM.MyFileGroup\\2.MyFile\\2 MSSQL:MyDatabase\\COM.MyFileGroup\\2.MyFile\\ 2 MSSQL$MyInstance:MyDatabase\.COM.MyFileGroup\ .2.MyFile\\2
MyDatabase\\COM MSSQL:MyDatabase\\COM MSSQL$MyInstance:MyDatabase\\COM
Table A-7
Command syntax for names containing a colon Name visible from SQL utilities SQL database named MyDatabase:COM. SQL filegroup named MyFileGroup:2 for the SQL database named MyDatabase:COM. SQL file named MyFile:2, which is a member of the SQL filegroup named MyFileGroup:2 for the SQL database named MyDatabase:COM.
Equivalent command-line Syntax

MyDatabase:COM MSSQL:MyDatabase:COM MSSQL$MyInstance:MyDatabase:COM MyDatabase:COM.MyFileGroup:2 MSSQL:MyDatabase:COM.MyFileGroup:2 MSSQL$MyInstance:MyDatabase:COM.MyFileG roup:2 MyDatabase:COM.MyFileGroup:2.MyFile:2 MSSQL:MyDatabase:COM.MyFileGroup:2.MyFi le:2 MSSQL$MyInstance:MyDatabase:COM.MyFileG roup:2.MyFile:2
A-20
Table A-8
Command syntax for names containing periods, back slashes, and colons Name visible from SQL utilities SQL filegroup named My/FileGroup.2 for the SQL database named My:Database.COM.

My:Database\.COM.My\\FileGroup\.2 MSSQL: My:Database \.COM.My\\FileGroup\.2 MSSQL$MyInstance: My:Database\.COM.My\\FileGroup\.2
A-21
A-22
Invisible Body Tag
B
This appendix describes how to use the NetWorker Module Permissions Wizard. This appendix includes the following main sections:

Permissions wizard features............................................................ B-2 How to run the permissions wizard............................................... B-2
B-1
Permissions wizard features

For release 4.0 or later, the NMSQL on Windows Server 2003 requires that the user have access to various system folders and resources. Because Microsoft Windows Server 2003 enforces new security constraints, the default user account is limited and does not have the necessary privileges to run the NMSQL. The permissions wizard, however, enables a local system administrator to easily configure user accounts and groups to run the NMSQL.
CAUTION If you are running the NMSQL on Windows Server 2003, make sure that you have the appropriate privileges before running a backup. If you do not have administrative privileges, you must run NMSQL from the console, not from Remote Desktop or terminal services. Otherwise it causes failures even if all permissions are correct.
How to run the permissions wizard

The Permissions Wizard runs automatically after installing or uninstalling the NMSQL software. For additional information on installing the NMSQL software, refer to the NetWorker Module for Microsoft SQL Server Installation Guide.
Note: When running the NMSQL User program in emulation mode on a SQL Server 2005 x64, the permissions wizard is deactivated. In this case, refer to the Release Notes for more information.
B-2
Additionally, you can run the permissions wizard from the NetWorker User for SQL Server program as shown in Figure B-1 on page B-3.
Figure B-1
Permissions tool option
To use the Permissions Tool: 1. Run the permissions wizard by using one of the following methods: Install the NMSQL software on Windows Server 2003. The permissions wizard runs automatically at the end of the installation process. From the NetWorker User for SQL Server program, select the Permissions Tool option from the Operation menu.
B-3
The Welcome to the NetWorker Module Permissions Wizard window appears.
Figure B-2
Welcome to the NetWorker Module Permissions Wizard window
2. Click Next to continue.
B-4
The NetWorker Module Selection dialog box appears.
Figure B-3
NMSQL Selection dialog box Note: For this initial version of the permissions wizard, the NetWorker Client list contains only one product, the NetWorker Module for Microsoft SQL Server.
3. Select Next to continue.
B-5
The Group and Permissions Administration dialog box appears.
Figure B-4
Group and Permissions Administration dialog box
4. In the Groups and Permissions Administration dialog box, select one of the following options and click Next: If you select the Create, delete, or edit a group option, continue with step 5. If you select the Verify and/or edit permissions of existing groups and/or users option, skip step 5 and continue with step 6. 5. From the Managing local or active directory groups dialog box, as shown in Figure B-5 on page B-7, select the appropriate radio button and click Next: Click the local group button to create, delete, or edit a local group This selection enables you to select the Local Group button to run the Windows Local Users and Groups application. From this window, you can create and manage users and groups
B-6
that are stored locally on the computer. Microsoft documentation provides additional information on using the Windows Local Users and Groups application.
Note: After managing users and groups, close the Local users and groups window to exit this Windows application.
Click the AD group button to create, delete, or edit an active directory group This selection enables you to select the AD Group button to run the Active Directory Users and Computers application. From this window, you can manage all Active Directory objects. Microsoft documentation provides For additional information on using the Active Directory Users and Computers application.
Note: After managing the Active Directory objects, close the Active Directory Users and Computers window to exit this Windows application.
Figure B-5
Managing Local or Active Directory Groups dialog box
B-7
The Selecting a user or group dialog box appears.
Figure B-6
Selecting a user or group dialog box
6. From the Selecting a user or group dialog box, select the appropriate option and click Next. Select a user or group to modify permissions To use this option, enter the name of a user or group in the text box, and then click the Check Name button. The permissions wizard verifies that the user or group you enter is valid. A name that passes this verification process appears underlined. Entering a name that is not valid returns an error dialog message. Display Object Permissions Required by the NetWorker Module software
B-8
The Verifying and Editing Permissions dialog box appears.
Figure B-7
Verifying and Editing Permissions dialog box
7. From the Verifying and Editing Permissions dialog box, follow the instructions to edit any objects that do not have the required permissions and click Next.
B-9
The Completion of Wizard dialog box appears.
Figure B-8
Completion of Wizard dialog box
8. From the Completion of Wizard dialog box, click Finish to close the wizard.
B-10
Invisible Body Tag
C
This appendix describes how the NetWorker Module for Microsoft SQL (NMSQL) uses striping to improve backup and recovery performance. This appendix includes the following main sections:

Striping with the NMSQL................................................................ C-2 Interleaving ........................................................................................ C-2 Striped backups................................................................................. C-3 Striped recoveries.............................................................................. C-5
C-1
Striping with the NMSQL

The NMSQL supports the use of multiple stripes for backing up and restoring Microsoft SQL Server 7.0, 2000, and 2005 data. Stripes are one or more streams of data that may be extracted, in parallel, from a database, and written in parallel to multiple media devices, such as tape drives. With the NMSQL, striping can yield a significant performance advantage when a large amount of data is backed up and restored by using multiple tape drives.
Note: Although SQL Server 2000 supports 64 stripes, the NMSQL supports only 32 stripes. This restriction applies to both the NetWorker User for SQL Server program, and the nsrsqlsv and nsrsqlrc commands.
You can specify striping from one of the following:
From the Backup Options dialog box in the NetWorker User for SQL Server program. From the NetWorker Administrator program for a scheduled backup. From the Windows command prompt on the client host. Append -Sn to the nsrsqlsv command, for example:
nsrsqlsv -s NetWorker_server_name -S3 db_name
where n is the number of stripes to use.

Note: Use the lowercase -s to specify the NetWorker server name and the uppercase -S to specify the number of stripes to use. Using the nsrsqlsv command on page A-3 provides more information about using the nsrsqlsv command.
Interleaving
Interleaving is the process of writing multiple stripes to the same volume. The NMSQL supports interleaving for backing up multiple clients to a single backup device. Recovering interleaved backups, however, can require substantial restore time. When restoring an entire interleaved backup, one stripe at a time is restored. This requires multiple passes over the same areas on the backup medium.
C-2
When restoring an interleaved backup, restore each clients backup separately.
Striped backups
Before any striped backup begins, the Parallelism attribute must be set to at least one more than the number of stripes being used. Set this attribute in the Set Up Server dialog box in the NetWorker Administrator program. For example, if you use three stripes, specify a value of four or greater for the parallelism.
Performance considerations for striping

For improved backup performance, follow these suggestions:
In the Create Device or Edit Device dialog box in the NetWorker Administrator program, configure target sessions for the NetWorker devices that are performing the striped backup. Always specify one session per device when you use striping. This yields the best backup performance on the client host by eliminating interleaving.
Allot a separate backup device (such as a tape drive) for each stripe in the backup or restore operation. Use a different NetWorker server for backup and restore operations not related to SQL Server. Dedicate a storage node exclusively to the striped backups.
The NMSQL automatically restores data by using the same number of stripes that were specified for the backup. For best restore performance, ensure that the same number of devices used for the backup are also available at restore time.
Configure the computer that runs the following : NMSQL software NetWorker client software SQL Server software as a NetWorker storage node
Note: The NetWorker server can be running on a different host.
Striped backups
C-3
For additional information about backups, see the following:

Chapter 2, Manual Backups Chapter 3, Scheduled Backups
Note: Chapter 4, Restoring SQL Server Data provides additional information about restoring backups.
Perform a striped backup

The NMSQL supports manual and scheduled backup striping. To perform a manual striped backup from the SQL Server host, which is a NetWorker client, use one of the following methods:
From the NetWorker User for SQL Server program: 1. From the Operation menu, select Backup. 2. Select an item to back up. 3. From the Options menu, select Backup Options. 4. In the Backup Options dialog box, check Create a Striped Backup. 5. Select the number of stripes from the Stripes list.
From the Windows command prompt, specify the -S option with the nsrsqlsv command. For example:
nsrsqlsv -sn NetWorker_server_name -Sn db_name
where n specifies the number of stripes you want to use, for example, -S3.
From the NetWorker 7.2 or earlier administrator program: 1. Select the appropriate Client resource. 2. In the Edit Client dialog box, select the Remote tab. 3. In Backup Command attribute, append -Sn to the nsrsqlsv command, where n is the number of stripes to use.
From the NetWorker 7.3 or later administration program: 1. From the Administration window, click Configuration. 2. In the expanded left pane, click Clients. 3. In the right pane, right-click the client, and select Properties.
C-4
4. In the Apps & Modules tab of the Properties dialog box, append -Sn to the nsrsqlsv command, where n is the number of stripes to use. 5. Click OK.
Striped recoveries
The NMSQL performs striped recovery optimizations so that striped recoveries to proceed as quickly as possible. This feature, represents a fundamental change in the striped recovery strategy, and impacts the capability to restore a striped backup under certain situations.
Optimal striped recovery versus fail-safe striped recovery

The NMSQL assumes a best-case scenario for striped recovery. The worst-case scenario cited in Performing a fail-safe striped recovery on page C-6 causes the striped recovery to stop responding by default. The restore process is unresponsive because the SQL Server is waiting for the NetWorker server to mount a volume. However, the NetWorker server has no tape sessions available. The NMSQL stops responding until the NetWorker server cancels the inactive tape session, which can take as long as 30 minutes. Table C-1 provides guidelines for determining when to activate the fail-safe striped recovery algorithm.
Table C-1
Guidelines for fail-safe striped recovery If The backup was interleaved Fewer devices are available Then Use the NetWorker Administrator program to determine if a second mount request has occurred for a volume that is already assigned tape sessions. Use the NetWorker Administrator program to determine if a device has been deactivated, or is busy performing an operation for another client.
Striped recoveries
C-5
Performing an optimal striped recovery (best-case scenario)

In a best-case scenario, striped recovery assumes, by default, that enough NetWorker tape sessions are available to enable the striped recovery to proceed without blocking. The NMSQL then assumes the following:
Each stripe was written to a different volume during backup. Interleaving was not used during the striped backup. The same number of devices that are used during backup are available during restore. No device failures have occurred since the backup. All devices that are used for backup are currently available. No other client sessions are currently assigned to the devices.
If the configuration does not meet these requirements, see Performing a fail-safe striped recovery on page C-6. If the configuration does meet all of these requirements, then the NMSQL optimally restores the striped recovery. The NMSQL begins delivering backup data to the SQL Server as soon as each tape session is established. This method provides the highest performance and lowest restore time. Earlier releases of the NMSQL permitted a striped recovery even if one or more of the requirements were not met. The NMSQL was required to load each volume that contain a stripe member before sending any backup data to the SQL Server. This method increased the restore time.
Performing a fail-safe striped recovery

If a striped backup is interleaved, or if fewer tape sessions are available at restore time (because of a failed or busy device), then a fail-safe striped recovery is necessary. The NMSQL determines the number of tape sessions available before starting striped recovery by contacting the NetWorker server with a list of striped save set. If the number of tape sessions is smaller than the number of striped save sets, the NMSQL uses a fail-safe striped recovery algorithm. The algorithm selected depends on the SQL Server version.
C-6
Note: For SQL Server 2000, the fail-safe algorithm might impact striped recovery.
Perform a Fail-Safe Striped Recovery for SQL Server 7.0 When enabling a fail-safe striped recovery to restore a SQL Server 7.0 client, additional configuration settings are not necessary.
Note: The following procedure applies to normal, copy, verify-only, and partial restore types.
To enable a fail-safe striped recovery on SQL Server 7.0: 1. In the Restore window of NetWorker User for SQL Server, mark the root item in the left pane. 2. From the Options menu, select Restore Options. 3. Deselect the Detect available tape sessions prior to restoring a striped backup attribute.
Note: This attribute is selected by default. It is controlled by the Windows registry entry NSR_DETECT_TAPES, which can be modified. This attribute retains its setting from one session to the next. Windows registry entry for striped backup on page C-10 provides more information on this keyword.
A dialog box, similar to Figure C-1, appears.
Figure C-1
Striped restore performance warning dialog box
4. Click OK in this dialog box, and then click OK to exit the Restore Options dialog box. 5. Start the recovery from the NetWorker User for SQL Server program or from the command line.
Striped recoveries
C-7
6. If the volumes with the striped recovery are not managed by an autochanger, then monitor events in the NetWorker Administrator program. When a media wait event occurs, load the appropriate volume.
Note: You might be required to load a volume more than once because of the fail-safe striped recovery algorithm implemented for SQL Server 7.0. For more information on this keyword, the NMSQL reads a required amount of information for each stripe into memory before proceeding.
7. Once the recovery is complete, clear the checkbox for the Detect available tape sessions prior to restoring a striped backup attribute in NetWorker User for SQL Server by following steps 1 and 2. Because this setting is maintained in the Windows registry, disabling the option enables the next striped recovery to proceed at maximum performance. Fail-safe striped recovery for SQL Server 2000 and 2005 When enabling a fail-safe striped recovery to restore a SQL Server 2000 or 2005 client, additional configuration settings are recommended to enable the restore to proceed at maximum throughput. The NMSQL takes advantage of the SQL Server 2000 and 2005 VDI feature called removable pipes. This feature allows third-party backup vendors to restore a striped backup from fewer devices. However, there is currently no way for the NMSQL to accurately determine how many tape sessions the NetWorker server can assign. The NetWorker server does not support striped recoveries by default. Therefore, only one tape session is available to restore the striped save sets during the predetection phase. Because of the way SQL Server 2000 and 2005 VDI removable pipes function, the third-party backup vendor can start only as many stripe restore threads as there are tape sessions available. Otherwise, the restore stops responding. Because of the removable pipes requirements and the NetWorker server limitation, the NetWorker Module, by default, restores one stripe save set at a time for SQL Server 2000 and 2005 when the Detect Available Tape Sessions Prior to Restoring a Striped Backup option is enabled. Therefore, if data was backed up with two stripes, the restore operation takes twice as long. You can, however,
C-8
temporarily enable striped recoveries on the NetWorker server to achieve maximum performance during a restore operation of SQL Server 2000 (or SQL Server 2005) striped backups.
Note: Because enabling striped recoveries on the NetWorker server may cause the file system restores to fail, do not use this recovery method as a permanent solution.
Perform a fail-safe striped recovery for SQL Server 2000 and 2005 The following procedure applies to normal, copy, verify-only, and partial (SQL Server 2000) or piecemeal (SQL Server 2005) restore types. To enable a fail-safe striped recovery on SQL Server 2000 or 2005: 1. In the Restore window of the NetWorker User for SQL Server program, mark the root item in the left pane. 2. From the Options menu, select Restore Options. 3. Select the Detect available tape sessions prior to restoring a striped backup attribute.
Note: This attribute is selected by default. It is controlled by the Windows registry entry NSR_DETECT_TAPES, which can be modified by either checking or unchecking this attribute. However it is set, this attribute retains its setting from one session to the next. Windows registry entry for striped backup on page C-10 provides more information on this keyword.
4. (Optional) Enable striped recoveries on the NetWorker server by creating the file striped_recovery in the ~\nsr\debug directory on the NetWorker server.
Note: You do not need to restart the NetWorker services to activate this setting.
5. Start the recovery from the NetWorker User for SQL Server program or from the command line. 6. If the volumes with the striped recovery are not managed by an autochanger, then monitor events in the NetWorker Administrator program. When a media wait event occurs, load the appropriate volume.
Striped recoveries
C-9
7. Once the recovery is complete, clear the checkbox for the Detect Available Tape Sessions Prior to Restoring a Striped Backup attribute in the NetWorker User for SQL Server program. Because this setting is maintained in the Windows registry, disabling the option allows the next striped recovery to proceed at maximum performance. 8. Once the recovery is complete, disable striped recovers on the NetWorker server by deleting the file striped_recovery in the ~\nsr\debug directory on the NetWorker server.
Windows registry entry for striped backup

The installation program for the NMSQL release 5.0 sets a Windows registry entry. This entry enables the Detect Available Tape Sessions option in the Restore Options dialog. By default, the entry is set to enabled. To change the default setting, select or clear the Detect Available Tape Sessions Prior to Restoring a Striped Backup attribute in the Restore Options dialog box. The current setting persists from session to session. You can also use regedit to change the default setting; modify the NSR_DETECT_TAPES entry in the following registry path:
HKEY_LOCAL_MACHINE\SOFTWARE\Legato\BSMSQL\Environment\
C-10
Glossary
This glossary contains terms related to disk storage subsystems. Many of these terms are used in this manual.
A
alternate track A track designated to contain data in place of a defective primary track. See also primary track. A set of access arms and their attached read/write heads, which move as an independent component within a head and disk assembly (HDA). Card that provides the physical interface between the director and disk devices (SCSI adapter), director and parallel channels (Bus & Tag adapter), director and serial channels (Serial adapter). Also known as a manual backup.
actuator
adapter
ad hoc backup
C
cache Random access electronic storage used to retain frequently used data for faster access by the channel. Unit of cache equivalent to one track. A database of information the NetWorker server maintains that tracks every database object, file, or file system backed up. The NetWorker server maintains a single client file index for each client computer.
g-1
cache slot Client File index
Glossary
Client resource
Identifies a NetWorker client, including data items such as file systems, disk volumes, directories, or files that will be backed up, and assigns specific backup instructions to the data. A data recording format employing self-defining record formats in which each record is represented by a count area that identifies the record and specifies its format, an optional key area that may be used to identify the data area contents, and a data area that contains the user data for the record. CKD can also refer to a set of channel commands that are accepted by a device that employs the CKD recording format.
Count-Key-Data (CKD)
D
DASD destage Direct access storage device. The asynchronous write of new or updated data from cache to disk device. A uniquely addressable part of the Symmetrix subsystem that consists of a set of access arms, the associated disk surfaces, and the electronic circuitry required to locate, read, and write data. See also volume. The hexadecimal value that uniquely defines a physical I/O device on a channel path in an MVS environment. See also unit address. The value that logically identifies a disk device in a string. A program used to initialize Symmetrix at installation and provide media maintenance.
device
device address
device number Device Support Facilities program (ICKDSF) diagnostics
System level tests or firmware designed to inspect, detect, and correct failing components. These tests are comprehensive and self-invoking.
E
EREP program The program that formats and prepares reports from the data contained in the Error Recording Data Set (ERDS). Enterprise Systems Connection.
ESCON
g-2
Glossary
F
Field Replaceable Unit (FRU) frame A component that is replaced or added by service personnel as a single entity. Data packet format in an ESCON environment.
G
Gigabyte (GB) Group resource 109 bytes. Names and defines a unique collection of Client resources that participate in a group backup operation. Clients can follow different backup schedules within the group operation, but all clients in the group start the operation at the same time.
H
Home Address (HA) The first field on a CKD track that identifies the track and defines its operational status. The home address is written after the index point on each track. The ability to define more than one logical volume on a single physical disk device making use of its full formatted capacity. These logical volumes are user-selectable in size. The minimum volume size is one cylinder and the maximum size depends on the disk device capacity and the emulation mode selected.
hyper-volume extension
I
ICKDSF identifier (ID) See Device Support Facilities program. A sequence of bits or characters that identifies a program, device, controller, or system. Initial microcode program loading. Indicates the physical beginning and end of a track. The reference point on a disk surface that determines the start of a track. An EMC-provided host-based Cache Reporter utility for viewing short and long term cache statistics at the system console.
g-3
IML index marker index point
INLINES
Glossary
instant backup
The process of creating a persistent snapshot (point-in-time copy) of a group of files, volume, or file system from a single client to a storage volume that can be recovered (used as a backup copy). The image is in native format, which is non-OTF. The instant backup is not backed up to tape, but its creation causes some metadata to be written to tape. The process of copying back the saved group of files or a file system from a mounted storage volume of a single client, which was created using instant backup. An addressable input/output unit, such as a disk device.
instant restore
I/O device
K
Kilobyte (K) 1024 bytes.
L
Least Recently Used Algorithm (LRU) logical volume The algorithm used to identify and make available the cache space by removing the least recently used data. A user-defined storage device. In the Model 5200, the user can define a physical disk device as one or two logical volumes. Requested data is not in cache and is not in the process of being fetched. Exclusive OR (XOR) of the accumulated bytes in the data record.
long miss
Longitude Redundancy Code (LRC)
M
manual backup A backup initiated by a sysadmin or operator as a one time, ad hoc procedure, as opposed to a scheduled backup initiated by the NetWorker server. The disk surface on which data is stored. A logical volume with all data recorded twice, once on each of two different physical devices.
media mirrored pair
g-4
Glossary
N
named log marks All transactions are logged for databases using the full recovery model. These transactions can be given unique names to simplify identification. The GUI program NetWorker administrators use to configure NetWorker Server resources and perform other administrative tasks on the NetWorker server. A storage device physically attached to another computer whose backup operations are administrated from the controlling NetWorker server.
NetWorker Administrator Program
NetWorker Storage Node
P
Partitioned Data Set (PDS) Assist An IBM feature for 3990 Model 6 and 3990 Model 3 with Extended Platform units. PDS Assist improves performance on large, heavily-used partitioned data sets by modifying the directory search process. Point in time. A fully usable copy of a defined collection of data, such as a consistent file system, database, or volume, which contains an image of the data as it appeared at a single point in time. A PIT copy is also called a snapshot. A NetWorker resource that specifies how long an entry is retained in the client file index. When a policy expires, the save sets associated with that policy are marked recyclable. A NetWorker resource that enables the administrator to direct backup data to selected volumes. A pool contains a collection of backup volumes to which specific data can be backed up. A NetWorker module that exports services of a storage subsystem by interfacing with vendor-specific APIs. This module is independent of applications and backup and recover interfaces. The original track on which data is stored. See also alternate track. The process of moving data from a track on the disk device to cache slot.
g-5
PIT Point-in-Time Copy
policy
pool
PowerSnap Module
primary track promotion
Glossary
Proxy Client
A surrogate client that acts as a remote data mover during snapshot operations. It performs the NetWorker save operation for the client that requested the backup.
R
recovery The process of restoring data to local storage on a client disk volume. The data can be recovered to the original source client, or by directed recovery to a different client. Data requested by the read operation is in cache. Data requested by the read operation is not in cache. The first record after the home address. The process of returning data to an earlier snapshot in response to a recovery operation. A rollback is a complete recovery of a storage subsystem from a point-in-time copy to a standard volume without host involvement. This may include an incremental recovery of changed blocks from a point-in-time copy to a standard volume in some snapshot technologies (such as TimeFinder) and switching a point-in-time copy with a standard volume (wherever applicable) in other snapshot technologies cases. As no host process is involved in copying data and data does not get retrieved from tape to disk, this type of recovery is faster in comparison to traditional recovery operations from offline media. rollover save set The save set that is backed up to tape from a snapshot. Whether this snapshot is retained or not is dependent on a snapshot policy. It is only when a snapshot is rolled over to tape that entries are made in the client file index and media database that permit these save sets to be browsed.
read hit read miss record zero rollback
S
savegrp A binary that is started when the time and interval specified in the Group resource matches. When a savegrp binary is started, it checks all the Client resources which have their group specified as the one which meets the time criteria and needs to be started.
g-6
Glossary
Save group is defined by a Group resource and run by the savegrp binary. save set A group of files or a file system from a single client computer backed up to storage media. An internal identification number that NetWorker software assigns to a save set. The process of reading, checking the error correction bits, and writing corrected data back to the source. For 3880 storage control emulations, this value uniquely identifies the storage director without respect to its selection address. It identifies to the host system, via the EREP, the director detecting the failing subsystem component. This value automatically increments by one for each director installed. The SCU_ID must be a unique number in the host system. It should be an even number and start on a zero boundary. A backup method that employs a proxy client to move the data from primary storage on the application server host computer to secondary storage. Requested data is not in cache, but is in the process of being fetched. The group of files, volumes, or file systems from a single client, describing the collection of data for which a point-in-time copy is created on an external disk subsystem, such as a storage array. A fully usable copy of a defined collection of data, such as a consistent file system, database, or volume, which contains an image of the data as it appeared at a single point in time. A snapshot is also called a point-in-time copy and can not be performed manually. For 3990 storage control emulations, this value identifies the physical components of a logical DASD subsystem. The SSID must be a unique number in the host system. It should be an even number and start on a zero boundary. The process of writing data from a disk device to cache. A series of connected disk devices sharing the same disk director.
save set ID
scrubbing
SCU_ID
serverless backup
short miss snap set
Snapshot
SSID
stage string
g-7
Glossary
striped backup
The SQL Server can generate multiple data streams that can be used to distribute a backup across multiple storage devices. For example a large database can be streamed to n storage devices, thereby shortening the backup time by approximately 1/n. Stripes refers to the concurrent threads of execution and backup data streams. For large databases, it may be desirable to store several data streams concurrently. Using more stripes than the number of available backup devices will result in interleaving and slow system operations considerably.
stripes
T
Traditional Backup A NetWorker backup performed as a scheduled process, as opposed to a Snapshot backup.
U
unit address The hexadecimal value that uniquely defines a physical I/O device on a channel path in an MVS environment. See also device address.
V
volume A unit of storage media, such as a magnetic tape or an optical disk. A storage device reads from and writes to volumes, which can be physical units (for example, a labeled tape cartridge) or logical units (for example, optical media can store multiple volumes on a single physical platter).
W
write hit write miss There is room in cache for the data presented by the write operation. There is no room in cache for the data presented by the write operation.
g-8
Index
A
action attribute 3-18 aliases attribute 3-13
B
backup backup device 2-11 canceling 2-12 combining SQL Server data objects 3-6 functions supported for SQL Server 6-5 group 3-10 in a cluster 5-15 levels comparison 3-5 levels, defined 1-11 levels, SQL Server terminology 1-11 manual 1-12 Microsoft Cluster Server data 5-3 monitoring 2-11 options volume pools 2-9 parallelism C-3 performance C-2 process overview, traditional 1-6 properties 2-10 required time 2-11 restrictions A-18 sample snapshot policy 3-9 scheduled 1-12 starting 2-11 striping C-4 transaction log backup 4-6
types supported for SQL Server 2000 6-4 backup command attribute A-3 backup options 2-7 backup status window 2-11 binary disk crash 7-6 BRC (Backup Recover Control) 1-10, 1-15 browse policy 3-11, 3-12 browse policy attribute 3-12 browse time, changing 4-33 bulk_logged recovery model 6-2
C
canceling a backup 2-12 canceling a restore 4-54 client resource 6-12, C-4 cluster virtual server 5-4 clustered SQL Server performing unscheduled operations on 5-15 Comments xii conventional backup 1-9 create time 1-19
D
data mover 1-11, 1-15, 4-24, 4-28 database file relocation restrictions 4-8 DBCC (database consistency check) 6-11 default group 3-10 default instance 6-13 default schedule 3-13 destructive restores overview 1-18 disaster recovery
i-1
Index
basic instructions 7-4 features 7-2 Rebuild Master utility 7-5 to 7-8 restore NetWorker binaries and online indexes 7-7 SQL Server 7-9 display conventions 1-21 distribution database 7-6 documentation, related x
M
marking items indicators 1-22 partially 1-23 restrictions 1-22, 1-24, 2-7 semantics 1-24 MSCS (Microsoft Cluster Server) active/active cluster configurations 5-4 active/passive cluster configurations 5-4 failover support 5-4 multiple instance 5-4 named instances 5-4 nodes supported 5-2 virtual server 5-2, 5-4 MSSQL save set A-16 multiple nonclustered instance 6-13
E
ERRORLOG file 4-2
F
failover cluster support 5-4 fail-safe recovery C-6 fake objects 1-20 filegroups backing up and restoring 4-8, 4-42 specifying for restore 4-43 files backing up and restoring 4-8, 4-42 full recovery model 6-2
N
named instances 5-4, 6-13, 6-14, A-17 named log mark 4-17, 4-49 defined 6-7 option 4-18, 4-51 specifying restore time properties 4-17, 4-49 NetWorker client file index 3-11 client resource 6-12, C-4 debug directory C-9 media database 3-11 NetWorker Remote Exec Service 4-2 NetWorker Administrator program backup schedules 3-14 configuring backup groups 3-11 display hidden attributes 3-8 starting a group 3-18 NetWorker client 3-11 NetWorker Module security requirements 1-2 support for MSCS clusters 5-3 NetWorker Module, interaction with 6-15 NetWorker User for SQL Server 1-23 1-24 Backup Status window 2-11 browse windows 1-20 display conventions 1-21 fake objects 1-20 marking items partially 1-23
G
group attribute 3-12
H
homogenous storage platforms 1-4
I
instant backup 1-8 instant restore 1-14 interleaving C-2
L
LAN and LAN-free environments 1-4 legacy database options select into/bulk copy 6-3 trunc.log.on.chkpt 6-3 logger (UNIX command) 3-18
i-2
Index
marking items restrictions 1-22, 1-24, 2-7 nwmssql command A-2 striped restore C-9 nonpersistent snapshot 1-9 NORECOVERY option 1-19, 4-7, 4-12 nsrd role during backup 1-13 nsrlog command 3-18 nsrmmd functionality during recover 1-13 nsrsqlrc command 4-1, 5-16, 6-14, A-2 command options A-9 failure 4-6 nsrsqlsv command 3-12, 5-13, 6-14, A-2, C-2 command options A-5 nwmssql command A-2 command options A-15
O
on-demand backup 2-2
P
parallelism C-3 partially marking items 1-23 password attribute 3-13 piecemeal restore 4-4 point-in-time 1-8 point-in-time backup 4-18 PowerSnap Module Backup Recover Control service 1-10 BRC (Backup Recover Control) 1-10, 1-15 PowerSnap Modules. See also snapshot 1-4 primary data disk crash 7-5 proxy client 1-9 proxy client. See data mover
R
Read File Configuration properties, specifying 4-47 Rebuild Master utility 7-6, 7-7 rebuildm command 7-5, 7-7, 7-9 recovery model
bulk_logged 6-2 changing 6-5, 6-7 full 6-2 simple 6-3 specifying 6-3 reducing network traffic 3-11 Remote Access attribute 3-13 Remote User attribute 3-13 removable pipes C-8 restore backup levels 3-5 canceling 4-54 change browse time 4-33 database file relocation restrictions 4-8 distribution database 7-6 ERRORLOG file 4-2 file destination 4-46 functions supported for SQL Server 6-5 in a cluster 5-15 master database 6-17, 7-6 model database 7-6 msdb database 6-17, 7-6 named log mark 4-17, 4-49 NetWorker binaries and online indexes 7-7 options 4-24, C-7, C-9 overwrite database 4-9, 4-54 overwriting existing database 4-14, 4-44, 7-3 point-in-time 1-19 process 4-7 properties 4-37 recovering file-configuration meta data 4-47 required time 4-9, 4-54 restrictions A-18 snapshot 1-12 snapshot methods 4-22 SQL cluster resources 6-16 status window 4-54 striped restore C-7, C-9 traditional recovery 1-12 transaction log backups 4-7 uncommitted transactions 4-7 using NetWorker User for SQL Server 4-8 viewing required volumes 4-35 wrong volume problem 4-9, 4-54 restore mode no recovery 1-19, 4-12 normal 1-18, 4-12, 4-13
i-3
Index
standby 1-19 restore time 1-19 database backup versions 1-20 named log marks 6-7 point-in-time 1-20 restore type copy 1-17, 4-25 normal 1-16 partial 1-16 verify only 1-17 retention policy 3-11, 3-12 Retention Policy attribute 3-12
S
Save Set atttribute 3-12 save sets 3-12 savegrp functionality 1-7 schedule attribute 3-12 scheduled backup backup strategies 3-4 creating backup groups 3-11 default group 3-10 levels 3-3, 3-4, 3-5 MSSQL save set A-16 NetWorker client 3-11 nsrlog command 3-18 requirements for Microsoft Cluster Server 5-5 schedule 3-13 snapshot policy 3-9 writing to log file 3-18 select into/bulk copy 6-2, 6-3 serverless backup 1-9 simple recovery model 6-3 snapshot backing up Microsoft cluster 5-14 backup 1-7 creating a snapshot policy 3-9 data mover 1-11, 1-15, 4-24, 4-28 homogeneous storage platforms 1-4 instant backup 1-8 LAN and LAN-free environments 1-4 restore 1-12, 1-14 severless backup 1-9 snapshot recovery operation 1-14
SQL Server backup and restore functions supported 6-5 binary disk crash 7-6 data syntax A-16 DBCC (database consistency check) 6-11 default instance 6-13 disaster recovery 7-9 distribution database 7-6 Enterprise Manager 6-11 instance 5-3 master database maintenance 6-11 model database 7-6 msdb database 7-6 named instances 6-13, 6-14 NORECOVERY option 4-7 primary data disk crash 7-5 Rebuild Master utility 7-5 to 7-8 reinstall, when to 7-4 services NetWorker Module interaction 6-15 production mode 6-15 single-user mode 4-6, 6-15, 6-17 virtual server 5-3, 5-4 SQL Server 2000 default instance 6-13 differential filegroup and file backups 1-5 fail-safe striped recovery C-9 multiple nonclustered instance 6-13, A-17 named log mark 4-17, 4-49 removable pipes C-8 running multiple nonclustered instances 6-13 SQL Server master database maintenance 6-11 STANDBY option 1-19 storage node attribute 3-12 striped backup interleaved C-6 interleaving C-2 manual C-4 parallelism C-3 performance considerations C-3 scheduled C-4 striped restore C-7, C-9 fail-safe striped recovery C-5 interleaving C-2 optimizations C-5 striped_recovery file C-9
i-4
Index
stripes defined C-2 fail-safe recovery on Microsoft SQL Server 2000 C-8, C-9 fail-safe recovery on Microsoft SQL Server 7.0 C-7 performing striped backup C-4 performing striped recovery C-6, C-7, C-8, C-9 performing striped recovery, optimal C-6 specifying a value C-3 syntax database, filegroup, and filename containing periods A-19
traditional recovery 1-12 transaction log backup 1-19, 4-6 maintenance 6-7 overflow prevention 6-8 transaction log backup 4-7 trunc.log.on.chkpt 6-2, 6-3
U
unmarking items indicators 1-22
V
virtual server 5-2, 5-3, 5-4 volume pools 2-9, 3-17
T
traditional backup 1-5
i-5
Index
i-6

MSSQL 5 Admin

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

MSSQL 5 Admin

Diunggah oleh

Hak Cipta:

Format Tersedia

EMC NetWorker Module for Microsoft SQL Server

EMC Corporation Corporate Headquarters: Hopkinton, MA 01748-9103

Copyright 1996 - 2006 EMC Corporation. All rights reserved.

Preface............................................................................................................................. ix Chapter 1 Overview

Restoring SQL Server Data

Backup and Recovery for Microsoft Cluster Server

Microsoft SQL Server

NetWorker Module Commands

NetWorker Module Permissions Wizard

Striped Backup and Recovery

Glossary ........................................................................................................................ g-1 Index ............................................................................................................................... i-1

Organization Chapters and appendixes in this guide are as follows:

Chapter 2, Manual Backups

EMC NetWorker PowerSnap Module Installation and Administrators Guide

Courier, italic Courier

User entry. For example:

Options in command line syntax

About the NetWorker Module for Microsoft SQL Server

NMSQL imposes the following requirements:

About the NetWorker Module for Microsoft SQL Server

NetWorker PowerSnap modules

Snapshot operations Serverless backup method

SQL Backup API

nsrindexd nsrmmdbd data

Traditional backup command and data flow

The NMSQL does not support:

The NMSQL supports three types of snapshot backups:

Instant Nonpersistent Serverless

NetWorker Client Microsoft SQL 2000 Server Databases and Logs

NetWorker Server File Index Media Database

nsrindexd PowerSnap nsrsqlsv BRC nsrmmdbd data nsrsnap nsrmmd

nsrexecd inter-process communication

Snapshot backup command and data flow

Where to initiate backup operations Backup levels available

Backup type Scheduled Manual

full Yes Yes Yes

incr Yes Yes No

diff Yes Yes No

NetWorker Client Microsoft SQL Server Databases and Logs

NetWorker Server File Index Media Database

SQL Backup API

nsrindexd nsrmmdbd data nsrmmd

inter-process communication nsrd

Traditional recovery command and data flow

nsrindexd PowerSnap nsrsqlrc BRC nsrmmdbd data nsrmmd inter-process communication

Snapshot recovery command and data flow

Rollback restore type

NetWorker User for SQL Server program

Marking a fake object

Left and right

Fake filegroup File

NetWorker User for SQL Server program

Partially marked items

NetWorker User for SQL Server program

Marking semantics and restrictions

The NMSQL imposes the following restrictions:

Restore window restrictions

Restore Options error message

Error logs for backup and recovery

nsrsqlsv.log nsrsqlrc.log xbsa.messages

About manual backups

Performing a backup with NetWorker User for SQL Server